Abstract
Clinical Oral Presentations
Abstract 0167 – ORAL Clinical
Helicopter and ground emergency medical services transportation to hospital after major trauma: A comparative cohort study
O Beaumont1, O Bouamra2, T Coats3,
1Oxford University Hospitals Trust
2Trauma Audit Research Network, University of Manchester
3Emergency Medicine Academic Group, University of Leicester
4Trauma Audit Research Network, University of Manchester
5Trauma Sciences, Blizard Institute, Queen Mary University of London
6University Hospital Coventry
7Research Centre, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Science, University of Oxford
A second analysis compared three similar patient cohorts admitted to both MTCs and TUs (n = 17,682). These were patients receiving either HEMS, GEMS with a Medical Emergency Response Incident Team (MERIT) or GEMS with standard crew taken to a TU but requiring time critical interventions (transferred to MTC within 48 h or died within 4 h). This analysis also employed multiple logistic regression and adjustment for centre effect.
For the secondary analysis, HEMS conferred a significantly reduced odds of mortality versus the GEMS cohort requiring time critical intervention admitted to a TU (OR: 0.73; 95%CI: 0.60–0.89). The GEMS+MERIT cohort showed a non-significant tendency to improved survival (OR: 0.88; 95%CI: 0.70–1.09).
Abstract 0179 – ORAL Clinical
Physician-documented pain assessment in 45 intensive care units in the United Kingdom. The PAin in INTensive Care (PAINT) study
C Bantel1, F Gordon2,
1Oldenburg University
2Imperial College London
P values for Mann Whitney comparisons or Spearman’s correlations for variables influencing the frequency of physician pain assessments.
Note: Those statistically significant are shown in bold.
Patient- and unit-specific characteristics showing a significant effect on number of physician-documented pain assessments.
The PAINT study was an immediately retrospective observational service evaluation of 45 ICUs in two trainee research networks in London and South East England. The protocol was registered at all sites (formal ethical approval deemed unnecessary; discussion with coordinating R&D department). Data collection occurred during one weekday 24-h study period. Medical and nursing records of all patients admitted to ICU aged over 18 years were scrutinised for any mention of pain, including use of pain assessment tools. Patient- and unit-specific factors were also collected. Univariate analyses of factors that may influence frequency of pain assessment were performed using Mann Whitney comparison or Spearman’s correlation. Factors identified as significant were tested using a hierarchical regression model (analysis with SPSS version 22).
Data were collected from 750 patient records reflecting the practice of 362 physicians. There was no physician documentation or mention of pain in 488 (65%) of patients’ notes and no behavioural pain assessment tools were used. Factors identified as influencing frequency of physician-led pain assessment are shown in Table 1 (univariate analysis) and Table 2 (hierarchical model). No nursing pain assessment was documented in 215 (29%) patients and only two ICUs used recommended behavioural pain scales.
This study shows that within a 24-h period nearly two-thirds of patients failed to have any mention of pain documented by physicians and suggests some of the factors that may influence frequency of assessment. Lack of documentation may not equate to the absence of an assessment but at best reflects a reduced prioritisation of pain. Pain assessment tools recommended by guidelines have not been adopted widely. Reasons for these findings are not clear but may be related to lack of knowledge, lack of time, perceived complexity or confusion surrounding assignment of duties within the ICU team. Future qualitative research is required to identify and address such factors.
References
Abstract 0210 – ORAL Clinical
Using historic trends to estimate the future prevalence and mortality of sepsis in Australian and New Zealand Intensive Care Units in 2025!
M Bailey1, S Huckson2, S McGloughlin3,
1School of Public Health and Preventive Medicine, Monash University
2The Australian and New Zealand Intensive Care Society (ANZICS) Centre for Outcome and Resource Evaluation (CORE)
3The Alfred Hospital
4The Australian and New Zealand Intensive Care Society (ANZICS) Centre for Outcome and Resource Evaluation (CORE)
Trends in mortality for patients admitted to ICU with sepsis 2000 to 2025.
Incidence, mortality and length of stay in ICU over the next four years closely matched trends extrapolated from the previous 12 years, with incidence rising to 11.4%, mortality declining to 15.6% and ICU length of stay declining to 67 h (IQR: 36–130).
Further extrapolation of trends into the future, suggested that by 2025, sepsis incidence might plateau at 11.5% of all ICU admissions and mortality might be expected to decline to 10% (Figure 1). Despite declining ICU length of stay, the projected increase in admission numbers to almost 25,000 per year would require the equivalent of a minimum of 65 additional ICU beds dedicated 100% of the time to sepsis admissions.
References
Abstract 0137 – ORAL Clinical
Neuron specific enolase analysis for neuroprognostication following out of hospital cardiac arrest: Test implementation and initial results from a district general hospital
1Intensive Care Unit, Royal United Hospital, Bath
2Honorary Professor of Resuscitation Medicine, University of Bristol
Impairment of neurological function is a key determinant of outcome after cardiac arrest. Predicting neurological recovery is difficult but remains fundamental in decisions about withdrawal of life-sustaining therapy (WLST). Neuroprognostication relies on clinical examination, neurophysiological studies, imaging and biochemical analysis. The 2015 Resuscitation Council (UK) post-resuscitation guidelines now advocate biochemical analysis.1 Neuron-specific enolase (NSE) is the most widely studied prognostic biomarker but is not in common use in the UK.
We started analysing NSE in May 2015 and have collected data on NSE values and patient outcomes for 22 patients admitted to our intensive care unit following cardiac arrest. To date, we have analysed 51 samples at a cost of £180 per test (including costs for NSE test kits, calibrators, and quality control processes). Each analysis kit can test 100 samples but requires replacement after one month. At absolute capacity, we estimate the cost per test could be as low as £9.22.
Introducing testing has been straightforward. We have educated clinical staff, (which has included the implementation of prognostication protocol) and have managed to ensure samples are sent at appropriate time intervals. Data sets were incomplete for some patients, but repeat samples were not sent if the patient had woken up or had WLST within the first 72 h. Two of 51 samples haemolysed during analysis but otherwise testing has been uncomplicated.
Of the 22 patients tested, eight survived to hospital discharge. We have grouped NSE values by cerebral performance category (CPC) and time (Figure).
Whilst our sample size is too small to draw any reliable conclusions our initial observations are consistent with those reported in larger studies2,3:
• High NSE absolute values and increasing values over time are associated with a poor outcome.
• Lower values are associated with neurological recovery, but some patients with low NSE values had evidence of significant hypoxic-ischaemic injury using other methods of neurological assessment.
• All eight patients with NSE values >100 µg/L died; however, a patient with a NSE value of 91 µg/L made a full neurological recovery.
Introducing NSE analysis has been straightforward and would be inexpensive if the laboratories undertaking the testing processed many patients (either high volume centres or by regionalising the service), thereby maximising use of each 100-sample kit. Further data are required to establish precisely how the measurement of NSE values after cardiac arrest contributes to multi-modal prognostication.
References
Abstract 0187 – ORAL Clinical
Impact of diagnostic venesection on haemoglobin levels in patients undergoing major cancer surgery during perioperative period
Royal Marsden Hospital
Patients undergoing major cancer surgery are particularly susceptible to perioperative anaemia. This arises from their underlying diagnosis and treatment, surgical blood loss and iatrogenic blood sampling.1,2 They often require allogenic blood transfusions to prevent developing anaemia.3 However, transfusions are associated with both harmful physiological consequences and significant cost. Hence, clinicians minimise the volume of red cell transfusions prescribed. As yet, no study has measured the actual amount of blood sampled from patients undergoing major cancer surgery in the perioperative period to determine the impact of this factor on perioperative anaemia. Our aim was to determine the total volume of blood sampled from patients for diagnostic purposes from pre assessment clinic until day of discharge.
We conducted a retrospective observational study at a specialist oncology hospital evaluating the total amount of blood venesected from patients undergoing elective major cancer surgery of any speciality associated with blood loss>300 ml. All blood samples were included in the study including blood gases (intra and postoperative) as well as miscellaneous tests such as serology. Blood bottles were estimated to hold a full volume. Discarded blood from arterial and central venous lines was estimated at 3 ml per gas.
In total, 86 patients were identified (49 male) with an average age of 62. The mean total blood venesected per patient during the perioperative period was 341 ml. We identified that the average length of admission was 20.6 days and the mean blood taken per day was 16.6 ml. The average haemoglobin drop was from 11.9 g/L from 115.1 gL to 103.2 gL, and the mean number of red cell units transfused was 4.2. The proportion of tests showed that blood gases accounted for 40% of all blood venesected.
This study is the first to highlight the magnitude of blood sampling in major cancer surgery during the entire perioperative period and identifies the high proportion that blood gas sampling in critical care contributes to perioperative anaemia. It emphasises the importance of avoiding unnecessary blood sampling and encourages the use of Patient Blood Management (PBM) concepts in diagnostic testing. This includes educating health professionals in PBM strategies such as the minimal sampling of blood and employing equipment that either avoids the need for repetitive venesection or the discarding of blood from sampling lines.
Mean amount of blood sampled per patient (341 ml) by department.
References
Abstract 0192 – ORAL Clinical
Clinical care and outcomes for ventilated bone marrow transplant recipients in a tertiary intensive care unit – A service re-evaluation
S Benington and L Howard
Central Manchester University Hospitals NHS Foundation Trust
Patients requiring mechanical ventilation following bone marrow transplantation (BMT) historically have a poor prognosis and are associated with therapeutic nihilism. We conducted an evaluation of processes of care and outcomes in this patient group in a large tertiary centre intensive care unit (ICU) and compared our performance with a previous evaluation of care for patients admitted between 2007 and 2010.1
Information was collected for mechanically ventilated bone marrow transplant patients admitted to our ICU between 2011 and 2015 with reference to case notes and prospectively collected admission ICNARC data. Results were compared with historical data from a previous service evaluation of 2007–2010 data.
A total of 483 transplants were performed 2011–2015 compared with 266 from 2007 to 2010.Of these, there were 44 episodes of mechanical ventilation (9%). APACHE II scores were similar between the two time cohorts. Patients were admitted between two days and two years post-transplant (median: 44 days), and median length of stay was 11 days. The commonest reasons for mechanical ventilation were: respiratory failure; septic shock; neurological issues. Compared with the previous time period, there was a reduction in lead time from first review to ICU admission, and a reduction in time from decision to admit to ICU admission. There were also fewer instances of multiple critical care reviews prior to admission. A significant increase in the incidence of appropriate key investigations was demonstrated (bronchoalveolar lavage (BAL), echocardiogram, CT thorax, viral PCR), although BAL was still infrequently performed. Haemofiltration was associated with increased ICU mortality (64% vs. 43%), but was much more likely to be associated with discharge alive from ICU than during the previous time period. No patient ventilated for neurological reasons was alive one year. Survival to ICU discharge was 50% (compared with 26% for 2007–2010), and one-year survival was 26% (compared with 6% 2007–2010).
Processes of care have improved following our initial service evaluation in 2007–2010. This has been accompanied by an improvement in ICU and one-year survival; it is difficult to determine how much is due to improvements in processes of ICU care or baseline differences in the transplant population and patient selection over time. Areas for improvement include a more robust approach to BAL sampling to aid timely diagnosis, and implementation of the BMT investigation checklist as a prompt in our new ICU electronic patient record system to further improve reliability of care.
Research Oral Presentations
Abstract 0215 – ORAL Research
An individual patient data analysis of tidal volumes used in three large randomised control trials involving patients with ARDS
R Lall1, C McDowell2, G Perkins1,
1Clinical Trials Unit, University of Warwick
2Northern Ireland Clinical Trials Unit, Queen's University
3Nuffield Department of Clinical Neurosciences, University of Oxford
4Regional Intensive Care Unit, Royal Victoria Hospital
Acute respiratory distress syndrome (ARDS) is a condition with a high mortality and morbidity,1 remaining static despite decades of familiarity.2 Mechanical ventilation prevents immediate mortality but may further damage patients’ lungs. Low tidal volume lung protective strategies have been shown to increase survival by reducing this iatrogenic damage. Current guidelines recommend tidal volumes of 6–8 ml kg−1 3 of ideal bodyweight.
We used data from three large randomised controlled trials of treatments for ARDS to determine compliance with these recommendations. We used the tidal volume recorded at randomisation for all patients in the OSCAR,4 HARP-25 and BALTI-26 studies. In addition, we used the ventilation data for control arm patients in OSCAR and all patients in HARP-2 at days 1 and 7 after randomisation.
The three trials enrolled 1660 patients from 46 intensive care centres across the United Kingdom, with tidal volume data available at least one time point in 1412 patients. Compliance with 6–8 ml kg−1 recommendation for tidal volume ranged from 19.7 to 38.5% of patients across all time points in all three trials.
Low compliance with the guidelines for tidal volume in patients with ARDS has been demonstrated before in case series, but not in clinical trials where the patient population is specifically selected against standard ARDS diagnostic criteria and the investigators were encouraged to use low tidal volumes. This study suggests that a review of the tidal volume guidelines is due as ventilation modes and sedation practices have changed since they were originally proposed and low tidal volume may not be practical in many patients. Furthermore, assessment of novel ARDS therapies may be impaired when compared with control groups who do not truly fulfil current gold standards of care.
References
Abstract 0241 – ORAL Research
Predicting escalation to intensive care in patients with suspected infection: Derivation of a clinical prediction rule: UPA3
B Dimitrov1, L Forni2, D Hargreaves3,
1University of Southampton
2Royal Surrey County Hospital FT
3Western Sussex Hospitals NHS FT
4University of Brighton
AUCROCs: UPA3: 0.81 (95% CI: 0.78–0.84), qSOFA: 0.58 (0.53–0.63), NEWS: 0.62 (0.57–0.66) and SIRS: 0.59 (0.54–0.63). AUCROC – area under the receiver operating characteristic curve.
Abstract 0242 – ORAL Research
External validation of the quick sequential organ failure assessment (qSOFA) and Systemic inflammatory response syndrome (SIRS) scores in a lower middle income country (LMIC) setting
K Gamage1, K Jayasuriya1, S Kariyawasam1, P Madushanka2, F Miskin2, S Premaratne1, A Weerarathne1, P Athapattu3, A Beane2, N De Silva1, P De Silva3, RMD Rathnayake1, C Sigera3, JA Sujeewa1, A Dondorp4,
1Monaragala District General Hospital
2Network for Improving Critical Care Systems and Training
3National Intensive Care Surveillance, Ministry of Health
4Mahidol Oxford research Unit
5Department of Clinical Medicine, University of Colombo
Early recognition of patients with sepsis is crucial in a LMIC, especially as critical care availability is limited. Availability of observations is an added challenge in this setting. qSOFA has been proposed as a superior to the previously SIRS-based sepsis definitions. This study is an external validation of qSOFA (AVPU was used instead of GCS) and SIRS (white cell count was not collected) in Moneragala District General Hospital (MDGH) in Sri Lanka.
A total of 15,577 consecutive adult (≥18 years) admissions from May to December 2015 were considered; 1844 admissions (11.8%) were due to infective causes as per ICD 10 coding and were included in this validation study. Observations from nursing charts and medical notes were extracted daily. Outcomes of interest were defined as deaths (20, 1.1%), ICU admissions (29, 1.6%), cardiac arrests needing CPR (30, 1.6%) and clinical transfers to a tertiary hospital (9, 0.5%). Sixty-seven (3.6%) patients experienced at least one of these events.
Observation availability is shown in Figure 1. Mean (SD) qSOFA score and SIRS score at admission were 0.58 (0.69) and 0.66 (0.79), respectively (Figure 1). Validity was assessed using area under the receiver operating curve (AUROC), Hosmer Lemeshow (HL) test and odds ratio over baseline risk (age) for the recommended qSOFA and SIRS cutoffs. qSOFA and SIRS both demonstrated poor discrimination for predicting events (AUROC = 0.63; 95% CI: 0.56–0.69 and AUROC = 0.62; 95% CI: 0.55–0.69, respectively) but were both well calibrated (HL statistic p = 0.51 and p = 0.27, respectively). AUROC for qSOFA and SIRS was not significantly different (p = 0.74). Discrimination for predicting deaths for qSOFA and SIRS was AUROC = 0.68 (95% CI: 0.55–0.82) and AUROC = 0.63 (95% CI: 0.50–0.76), respectively, with HL statistic of p = 0.16 and p = 0.046, respectively. AUROC values for deaths were also not significantly different (p = 0.31). Odds ratios over baseline risk (age) for qSOFA and SIRS are illustrated in Figure 2.
This first validation study of qSOFA in a low acuity DGH in an LMIC demonstrates poor discrimination and good calibration in predicting adverse outcomes at admission for hospitalized patients with infections but is overall no better than the previous SIRS criteria. Observation availability (especially AVPU) needs to be improved.
Data availability of qSOFA and SIRS variables and scores distribution. Odds ratios across baseline risk percentiles for qSOFA and SIRS.
Abstract 0050 – ORAL Research
A Biological effect or an Imbalance of co-interventions? Exploratory analyses of the Eurotherm3235 trial
PJD Andrews1, C Battison1, P Hopkins2,
1The University of Edinburgh
2King's College Hospital
3Université Libre Bruxelles
4Erasmus Center, Rotterdam
In this trial, titrated therapeutic hypothermia (TH) was used as the primary intervention to reduce elevated ICP. The effect of this intervention on outcome was assessed using the Glasgow Outcome Scale Extended (GOSE) at six months. The adjusted common OR for GOSE was 1.53 (95% confidence interval: 1.02 to 2.30, p = 0.04) indicating worse functional outcomes in the hypothermia group than in control group.1
The data were gathered on Case Report Forms (CRFs) from 1 January 2016 until 31 July 2016.
For all data collected, descriptive statistics by allocated treatment and dichotomised by GOSE outcome (favourable vs. unfavourable) were generated.
The most common reasons for death were multiorgan failure, infection (sepsis and pneumonia) as well as acute respiratory distress syndrome (ARDS) (33%) in the TH group, while brain stem death was most frequent (33%) in the CG.
The number of patients undergoing decompressive craniectomy, before and after randomisation, was similar in TH and in the CG (n = 48 and 53); the usage of hemi-craniectomy was slightly higher in the control group (∼10% more).
Two hundred and thirty of the 262 patients received hypertonic treatment for raised ICP (n = 113 TH, 117 CG).
Propofol usage was similar in both groups despite guidance to reduce rates after induction of hypothermia.
Two hundred and seventy-eight CRFs have been received and will be analysed before the SOA.
Figure 1. Cause of Death
Reference
Abstract 0103 – ORAL Research
Prediction of 30-day mortality in patients with sepsis: An exploratory analysis of care process and patient characteristics
E Blyth1, M Chikhani2, T McKeever3, I Moppett2,
1Chesterfield Royal Hospital NHS Foundation Trust
2Academic Anaesthesia and Critical Care, Division of Clinical Neuroscience, University of Nottingham
3Division of Public Health and Epidemiology, University of Nottingham
4School of Medicine, University of Nottingham
5Nottingham University Hospitals NHS Trust
Sepsis represents a significant public health burden, costing the National Health Service an estimated £2.5 billion annually, with 35% one-year mortality.1 The aim of this exploratory study was to investigate risk factors predicting 30-day mortality amongst adults with severe sepsis based on previous definitions2 and equivalent to sepsis using the current definition.3
Data were collected on patients with severe sepsis admitted to critical areas in our institution between November 2011 and March 2014 as part of our rolling quality assurance programme. Data were collected by dedicated multidisciplinary and multispecialty staff and included patient characteristics, components of SIRS (systemic inflammatory response syndrome), organ dysfunction, tissue hypoperfusion, location of early care provision, source of sepsis and timeliness of care interventions. These 97 separate variables were then investigated to determine which best predicted 30-day mortality. Data had previously been cleaned as part of on-going database maintenance, and were analysed with multivariate logistic regression using p < 0.05 on univariate analysis for initial inclusion and p > 0.05 for exclusion from the model.
Multivariate logistic regression model of variables significantly associated with 30-day mortality.
Odds ratio.
95% confidence interval.
Persistent systolic blood pressure <90 mmHg or mean arterial pressure <70 mmHg despite fluid resuscitation.
Data missing from 15 patients.
Adjusted odds ratio.
These data confirm some previous findings and also raise some new questions. Increasing age, thrombocytopenia, remaining hypotensive after vasopressor administration, increased serum-lactate concentration and mottled skin were all associated with greater 30-day mortality. The finding that hospital-acquired sepsis is a risk factor for mortality is consistent with some studies. From our data, it is not possible to say whether the protective effect of diagnosis on a surgical ward is due to a difference in pathology, early recognition or an eminently controllable source. Similarly, fever – a cardinal sign of sepsis – may result in more prompt recognition or be protective in its own right. These are worth exploring in further detail with respect to human factors involved in diagnosing and managing sepsis, specifically with respect to the surprise finding of treatment timeliness in this study.
References
Abstract 0124 – ORAL Research
A national study of long-term outcomes of patients admitted to Scottish intensive care units with chronic obstructive pulmonary disease
1Department of Anaesthesia, Critical Care and Pain Medicine, Royal Infirmary of Edinburgh
2Usher Institute of Population Health Sciences and Informatics, University of Edinburgh
Patients with chronic obstructive pulmonary disease (COPD) are frequently admitted to hospital. Previous studies have shown that doctors are reluctant to start life support due to concerns that outcomes for patients are poor.1 However, little research has been published world-wide relating to long-term consequences of ICU admission.2 Using the national database of ICU admissions, our objectives were to report one-year mortality and emergency hospital readmission rate in those who survived their index ICU admission and to identify factors associated with mortality and risk of emergency readmission.
We used a complete, national nine-year cohort of patients admitted with COPD to Scottish ICUs (01 January 2005–31 December 2013) from the Scottish Intensive Care Society Group (SICSAG) database, linked to hospital and death records. NHS ethical approval was waived. Study population: aged ≥35 years; primary diagnosis of COPD or primary diagnosis of pneumonia with COPD secondary diagnosis; patients who survived index admission.
There were 1753 patients with COPD exacerbations or pneumonia with COPD admitted to Scottish ICUs who survived their index admission (incidence 6.4/100,000 population). Mortality at 30 days was 2.1% (95%CI: 1.4, 2.8%) and at one year was 17.9% (16.0, 19.7%). Median time to death was 166 days (IQR: 71,252). The proportion readmitted within 30 days was 11.1% (95%CI: 9.7, 12.6%) and within one year was 52.2% (49.9, 54.5%). The median time to first readmission was 81 days (IQR: 23,182). The total number of emergency readmissions was 318 during follow-up, equating to 1.44 (95%CI: 1.38, 1.50) readmissions per person-year. Readmission rate was highest in month one of follow-up (2.20, 95%CI: 1.97, 2.46) and declined to a steady rate during months 7–12 (1.24, 95%CI: 1.06, 1.46). Statistically significant independent predictors of one-year emergency readmission were: number of emergency hospital admissions in the year before ICU admission, total number of comorbidities, source of ICU referral, acute physiology score and PaO2:FiO2 ratio. Mechanically ventilated patients were less likely to die (p = 0.003) and less likely to be readmitted compared to non-ventilated patients (p < 0.001) (Figure 1).
Our study demonstrated that COPD patients who survive their ICU admission have a substantial readmission rate in the year after hospital discharge. This is important information for clinicians and patients when considering post-ICU care and for policy makers in order to plan healthcare resources. Figure 1 – Readmissions MV vs. non-MV groups.
Kaplan-Meier plot for one-year post-discharge emergency hospital readmission for mechanically ventilated and non-ventilated COPD patients (PTO).
References
E-Posters
Abstract 0189 – E-Poster EPM.001
Exploratory analysis on the need for an ECMO eCPR service in South East London
G Auzinger1, T Best1, S Butt1, E Gelandt1,
1King's College Hospital NHS Foundation Trust
2London Ambulance Service
Survival after refractory cardiac arrest is poor and ECMO (eCPR) has been demonstrated to be superior to conventional CPR in patients suffering refractory cardiac arrest.1
King’s College Hospital is a major acute emergency, tertiary and cardiac arrest centre which receives such patients, and an ECMO service is available onsite providing a comprehensive ECLS programme with excellent outcomes.2,3
eCPR has been undertaken for a number of in and out of hospital arrests – survival in those with true eCPR is 43% (n = 21), and 50% in those with either true eCPR, an OHCA at presentation, or when ECMO was initiated in ED for those with transient ROSC only (n = 28).
We sought to establish the need for a more formal service to be developed as many patients with refractory cardiac arrest are not transported to hospital, but rather they undergo full ALS in the community supported by a tiered ambulance service response.
• Age 16–65
• Bystander CPR but no ROSC at 20 min
• Within a 20-min drive time of King’s College Hospital
• Monday to Friday 08:00–18:00
• Emergency calls allocated the highest priority response from the ambulance service
Patients suffering traumatic cardiac arrest were excluded.
Both patients transported to King’s with CPR ongoing and those where paramedics discontinued resuscitation and undertook a ‘recognition of life extinct’ protocol were included.
There were 21 patients (median age: 53, range: 24–62) conveyed to hospital with a median scene time of 46 min and 37 patients (median age: 49, range: 19–65) not transported – 72% were presumed cardiac in origin.
There should be a nationally led assessment of the need for eCPR within the UK and support for local providers innovating in this area.
References
Abstract 0212 – E-Poster EPM.002
Experience with a veno-venous extra corporeal CO2 removal device: Easy and safe, but only modest improvement in gas exchange
G Auzinger
King's College Hospital NHS Foundation Trust
References
Abstract 0073 – E-Poster EMP.003
An audit of intracranial bleeding in Extracorporeal Membrane Oxygenation
R Basra1, J McLean2, R Porter1,
1Glenfield Hospital
2London North West Healthcare NHS Trust
3Chelsea and Westminster Hospital NHS Foundation Trust
4Imperial College School of Medicine
References
Abstract 0205 – E-Poster EMP.004
A modern ICU dilemma: Should some patients be considered for immediate extracorporeal support without trial of invasive ventilation?
St George's Hospital
Extracorporeal membrane oxygenation (ECMO) is indicated in management of severe reversible respiratory failure. It is used primarily as a rescue treatment when conventional ventilation has failed.1 As ECMO improves and becomes increasingly available with advances in technology and practice, there is an argument it could be considered first-line therapy in certain cases. It has been trialled in awake, self-ventilating patients as a bridge for lung transplantation with positive results.2 Some centres are considering ECMO as a first-line treatment in acute respiratory failure, but there is currently a lack of supporting clinical evidence.3
A 23-year-old male was admitted with isolated chest trauma following a road traffic accident. CT scan revealed multiple rib fractures, bilateral lung contusions, pulmonary haemorrhage and a large right-sided lung cavity of unknown chronicity or aetiology (Figure 1).
Bilateral thoracotomies and chest drains were performed pre-hospital, and acute surgical intervention was not indicated.
Despite non-invasive ventilation and FiO2 1.0, he remained severely hypoxic with a P/F ratio of 5.5 kPa. Invasive ventilation was considered high risk due to his underlying lung pathology and concern regarding possible shunting and V/Q mismatch, or causing barotrauma to his cavitating lesion.
Discussions with the ECMO team focused on the risk and benefit of intubation prior to ECMO. Ultimately, the decision was made to trial invasive ventilation but, after intubation, peripheral saturations dropped immediately to 40%–50% on FiO2 1.0. Recruitment manoeuvres and proning were unsuccessful with SatO2 recoverable to only 70%.
At that point, his Murray Score was 3 and the decision was to proceed to VV-ECMO.
After 3 h, the patient was established on ECMO and was transferred to the ECMO unit. He remained on ECMO for seven days in total and was extubated two days after decannulation. He was repatriated to his local hospital, where unfortunately, he absconded.
On reviewing the case, discussion centred on the dilemma of invasive ventilation prior to extracorporeal support. The patient suffered a prolonged hypoxia after intubation that could have caused harm and, potentially, been avoided but the complexity and difficulties in managing patients on ‘awake ECMO’ provide a major challenge, with limited clinical data in the acute setting.3
Given the improvements in safety and access to ECMO, we propose that further research should be undertaken to establish ECMO as a potential first-line management of acute, reversible respiratory failure.
References
Abstract 0162 – E-Poster EPM.005
Use of Impella.5 and veno-arterial extracorporeal life support (VA ECLS) in severe cardiogenic shock secondary to refractory arrhythmia
Y Abdeldaim1, E Carton2, S Duff2, T Kiernan1, J McNamara2 and C Nix1
1University Hospital Limerick
2Mater Misericordiae University Hospital
Patients with cardiogenic shock secondary to refractory ventricular tachycardia (VT) have a poor outcome. Catheter ablation is the second-line therapy but is frequently impaired or prohibited by patients’ haemodynamic instability.1–3
A 46-year-old female patient presented with cardiogenic shock secondary to recurrent episodes of polymorphic VT and ventricular fibrillation (VF). Three years previously, she had presented with atrial fibrillation which progressed to polymorphic VT after administration of metoprolol. Following successful cardioversion, diagnostic cardiac investigations were reported as normal. An implantable cardioverter defibrillator (ICD) was inserted at that time.
On this admission, the patient developed cardiogenic shock secondary to recurrent episodes of VT/VF and the multiple ICD and external shocks delivered. On interrogation of the ICD, 60 shocks had been delivered in the previous 10 h. The patients required tracheal intubation, mechanical ventilation, high-dose vasopressor infusions and prolonged periods of resuscitation with a mechanical chest compression device (Lucas, Physio-Control). Multiple antiarrhythmic agents, including flecanide were administered. Mechanical circulatory support was achieved with a catheter-based temporary ventricular assist device (Impella 2.5, Abiomed) inserted during active chest compressions. The number of episodes of VT/VF requiring defibrillation decreased over the following 12 h; however, the patient remained in cardiogenic shock with persistently high plasma lactate concentration and anuric renal failure.
The ECLS service of our center was contacted and a decision was made to send a retrieval team to the referral hospital to initiate peripheral VA ECLS (Cardiohelp, Maquet). Percutaneous arterial access was not possible and a surgical cut down on the left femoral artery. Satisfactory ECLS blood flow was achieved and the Impella device was removed prior to transfer. There was a rapid improvement in the shock state and the vasoactive infusions were weaned. While still on VA ECLS, an arrhythmogenic focus was located and successfully ablated. VA ECLS was weaned on day 7. The patient made a complete neurological recovery, and no further episodes of VT/VF were recorded. The patient was discharged home and resumed full employment three months after presentation.
Use of the Impella 2.5 device either alone or combined with peripheral VA ECLS should be considered early in cardiogenic shock patients with potentially reversible recurrent ventricular tachycardias. A sufficient period of haemodynamic stability has to be provided to locate and treat the cause of the arrhythmia. The improved haemodynamic support is also likely to reduce the chances of coronary or cerebral ischaemic injury.
References
Abstract 0184 – E-Poster EPM.006
Virtual critical care follow-up
M Krupa, P O'Shea,
St George's Hospital
Critical care admissions can be associated with physical, cognitive and psychological sequelae after discharge. This is known as ‘post intensive care syndrome.’1 In 2009, the National Institute for Clinical Excellence introduced guidance that critical care patients should be followed up two to three months after discharge.2 However, a study in 2013 found, despite an increased awareness of need for follow-up, there was a low prevalence of services available.3 Reasons for this included lack of resources and prioritisation to other clinical areas.3
Our institution runs an outpatient follow-up clinic for patients discharged from intensive care unit (ICU). We recently piloted a virtual follow-up clinic, with the aim to reduce did not attend (DNA) rates.
The virtual clinic consists of a structured phone call by a critical care nurse. We created a questionnaire to enable a qualitative assessment of their symptoms. This includes asking patients about their memories of ICU and assessing their physical, psychological and functional status. If they report significant symptoms, they are invited to clinic. If they do not have any symptoms or concerns they are discharged.
We ran a pilot study from July to September 2016, which included 26 patients.
The two patients scheduled for outpatient clinic were seen outside of standard clinic times, due to conflicting health appointments (e.g. dialysis). We were able to avoid 11 DNA’s, including patients who were unable to be contacted and those not interested.
The virtual clinic is a quick tool to highlight patients who may be suffering from post-intensive care syndrome and to triage how urgently patients need to be seen. It can also identify patients who may be experiencing symptoms but did not fully meet outpatient criteria. It allows more flexibility for patients as clinic could be rearranged, or they could be followed up by phone only or referred to their GP, if patients preferred. Discharging patients who did not have concerns or complications allows other patients in need the opportunity to be seen earlier. Additionally, it can reduce DNA rates and shorten wait times.
This pilot study had a small patient cohort as limited resources reduced the amount of patients that could be contacted. A more extensive study with a larger cohort is required.
Criteria for invitation to outpatient follow-up clinic and virtual follow-up clinic.
Outcomes of patients contacted via virtual follow-up clinic.
References
Abstract 0153 – E-Poster EPM.007
The safety of mobilisation on inotropes
C Pereira
Royal Brompton and Harefield NHS Foundation Trust
The importance of early mobilisation to combat the effects of critical illness has been well documented. One factor reported to limit early physical rehabilitation is the presence of continuous infusions of inotropes/vasopressors. There is a paucity of data regarding this practice to guide clinicians considering physical activity with patients on vasoactive therapy.
Abstract 0223 – E-Poster EPM.008
Rehabilitation practices within three critical care units in the UK – A service evaluation
University Hospitals Coventry and Warwickshire
References
Abstract 0230 – E-Poster EPM.009
Patient-centered care or a step too far
L Enoch, M Georgieva and
JVF Critical Care Unit, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust
Early rehabilitation reduces physical impairment and decreases ICU length of stay.1 Rehabilitation should be individualized, structured and consistent. It requires multi-disciplinary involvement and in some cases alternative approach. We are presenting a challenging case of a critically ill patient for whom a trip home while still ventilated was the moving power towards successful weaning from mechanical ventilation and hospital discharge.
A 79-year-old patient was admitted in our ICU with severe pancreatitis. His stay was complicated with multiorgan failure and recurrent chest infections. After 120 days on the unit the acute inflammatory issues settled, the pancreatitis resolved. The patient was looked after our multidisciplinary team. He received individualized physiotherapy programme since admission and his nutrition was optimized on a daily basis. The main remaining issues were low mood and lack of progress with weaning from mechanical ventilation. The patient was regularly visited by his family, was engaged with areas of interest as watching cooking programmes and going for a trip to the hospital garden. The patient was a farmer and was actively involved with gardening. He expressed a desire to go home for one afternoon. The idea was explored during a multidisciplinary team meeting. A careful planning was initiated. The legal aspects were discussed with the trust lawyer. Risk assessment was done acknowledging home environment, staff and equipment needed. Discussion with the patient and his family regarding their expectations, and the plan for the trip was performed. The funding was provided by the charitable fund of our ICU.
One week later, we took the ventilated patient home. He spent 3 h with his extended family, neighbors and close friends. It was a very emotional experience for the patient and all our team.
From the following day, the mood of the patient improved significantly; he was highly motivated and got engaged with the intense physiotherapy. Eight days later, the patient was decannulated. Subsequently he was discharged from hospital 12 days later. The patient was followed up for the next six months. He remained highly spirited, his physical activity improved and there were no new chest infections.
Early rehabilitation is a key for optimal management of the critically ill patient. It requires a multidisciplinary approach and individualized planning. Patient-centered care is not a step too far; it is the step towards better quality of care.
Reference
Abstract 0101 – E-Poster EPM.010
The role of the Band 4 physiotherapy assistant practitioner in intensive care
Papworth Hospital NHS Foundation Trust
Papworth Hospital NHS Foundation Trust is one of the largest specialist cardiothoracic hospitals in Europe, and includes England’s main heart and lung transplant centres. The Band 4 physiotherapy assistant practitioner (AP) role was implemented in 2010. It was created as part of the Trust’s programme to develop a larger pool of unregistered staff within nursing and allied health professionals. Two AP posts were based on the cardiothoracic surgery wards, treating routine patients from day two to discharge; promoting independent mobility, performing stair assessments and providing discharge information. The AP’s also assist qualified physiotherapists with long-term slow rehabilitation patients requiring two staff during treatment sessions.
The APs staff the physiotherapy weekend rehabilitation service. Therefore, patients requiring rehabilitation receive input seven days per week; this ensures continued progression in physical improvement.
All APs complete basic competencies over the first three months of employment to ensure they receive adequate training before starting weekend working. The APs rotate between teams every 16 weeks and continue to increase experience and training on each rotation. They attend department and team training sessions, in addition to having personal mentors and line managers to provide additional support. The APs are also encouraged to attend appropriate external courses to increase their skills and knowledge.
Within ICU, the APs assist the qualified staff by providing additional therapeutic support during rehabilitation session with debilitated patients, i.e. sitting on the edge of the bed and providing independent additional rehabilitation treatments sessions of the day sessions such as limb exercises, bed bike or chair pedals, standing practice or marching on the spot. The APs treat the full range of conditions found within ICU; these include patients following cardiothoracic surgery, cardiology, heart failure, ventricular assist devices, post-transplant and patients receiving extracorporeal membrane oxygenation. These patients may be mechanically ventilated via tracheostomy, or self-ventilating with or without non-invasive ventilator support.
Recent developments include the APs undertaking screening to assess patients for treatment requirement day 1 post-operative cardiac surgery. This involves reviewing observations, blood gases and chest X-rays. There are also plans for collaborative working with the occupational therapists to increase cognitive rehabilitation with ICU.
Abstract 0013 – E-Poster EPM.011
The modulation of neutrophil extracellular traps in chronic critical illness: Results of a preliminary translational sub-study
D Dosanjh1,
1Institute for Inflammation and Ageing, University of Birmingham
2Department of Critical Care Medicine, Queen Elizabeth Hospital
Abstract 0149 – E-Poster EPM.012
GRiP: Does GM-CSF restore neutrophil phagocytosis in critically ill patients?
S Baudouin1, S Bowett2, T Chadwick3, A Conway Morris4, P Corris5, T Fouweather3, T Hellyer5, V Linnett6, J Macfarlane5, J Parker2, E Pinder5,
1Intensive Care Unit, Royal Victoria Infirmary
2Newcastle Clinical Trials Unit, Newcastle University
3Institute of Health and Society, Newcastle University
4Department of Anaesthesia, University of Cambridge
5Institute of Cellular Medicine, Newcastle University
6Intensive Care Unit, Queen Elizabeth Hospital
7Integrated Critical Care Unit, City Hospitals Sunderland
8Intensive Care Unit, Freeman Hospital
9Regional Intensive Care Unit, Royal Victoria Hospital
This study tested the hypothesis that subcutaneous GM-CSF restores neutrophil phagocytosis in critically ill patients. A beneficial effect of GM-CSF would potentially lay the foundations for an immunomodulatory strategy to reduce the incidence of nosocomial infection in critically ill patients without antibiotic intervention.
Patients were randomized to receive either subcutaneous GM-CSF (3 µg/kg/day) or placebo once daily for four days. Blood was drawn at baseline and on alternate days following initiation of the drug/placebo.
The primary outcome was neutrophil phagocytic capacity on day 2 following initiation of the administration of GM-CSF/placebo. Secondary outcomes included serial assessment of neutrophil phagocytosis, monocyte HLA-DR expression and safety parameters. Clinical outcomes such as mortality, ICU length of stay, duration of mechanical ventilation, PaO2 to FiO2 ratio and episodes of ICU-acquired infection were observed.
Abstract 0218 – E-Poster EPM.013
Oncostatin M (OSM) mediates inflammation in clinically relevant models of the acute respiratory distress syndrome (ARDS)
M Fitzgerald, D McAuley and C O'Kane
Centre for Experimental Medicine, Queen's University
Oncostatin M (OSM) is an IL-6 family cytokine recently recognized to be upregulated in the lung in acute respiratory distress syndrome (ARDS). In other pulmonary conditions, OSM synergises with other inflammatory mediators to drive a pro-inflammatory response, but the role of OSM in ARDS is uncertain.1
We hypothesised that OSM mediates inflammation in pulmonary endothelial cells, and that OSM neutralisation in in vitro and ex vivo models decreases pulmonary chemokine and protease production.
Stimulation with OSM/TNF-α (10 ng/ml) on chemokine and protease/anti-protease production.
Note: Data are mean (SEM).
White cell count (WCC) in bronchoalveolar lavage fluid at 4 h after injury with LPS and concurrent instillation of OSM inhibitory antibodies (OSMi) or isotype control.
In the human EVLP model, OSM neutralisation was associated with decreased leucocyte infiltration (Figure 1). Lungs showed less severe histological injury, as assessed by a composite score of intra-alveolar haemorrhage, alveolar protein and septal oedema, from mean (SEM) 7.0. (0.3) to 4.8 (0.5) (p = 0.03).
OSM inhibition is a candidate therapy for ARDS and should be further investigated in clinical studies.
References
Abstract 0204 – E-Poster EPM.014
Neurally adjusted ventilatory assist in prolonged mechanical ventilation: The challenges of conducting a feasibility trial of a complex intervention
1King's College London
2King's College Hospital NHS Foundation Trust
Preliminary data.
Note: Data are median (IQR) or number (percentage).
APACHE II: Acute Physiology and Chronic Health Evaluation
Patients receiving >2 h in a weaning ventilation mode within 28 days from randomisation.
In order to extract the maximum value from the trial, the study design has been adapted to ensure compliance with recent guidance.3 The aims and intervention are unaltered, but the changes do provide challenges in relation to the usual apriori outcome definitions, analyses plans and sample size calculations. These issues are not unique to this study and we hope that open, transparent discussion will be of use to others.
References
Abstract 0105 – E-Poster EPM.015
Greater reductions in monocyte HLA-DR expression precedes nosocomial infection in patients requiring critical care admission following major surgery
C Hinds1, B Lalabekyan2, R Longbottom2, M O'Dwyer1, R Pearse1 and H Torrance1
1Queen Mary University of London
2Barts Health NHS Trust
Demographic and clinical features of the study cohort.
Note: Data are presented as absolute counts and in parentheses either the interquartile range or a percentage. The percentage presented in the infection columns represents the percentage of total patients either with or without infection that possess the variable of interest. For example, 57% of patients who develop a post-operative infection were male in comparison with 37% of patients who did not develop a post-operative infection. The number in the total column may not always equal the combined value of the number in the subsequent two columns due to missing data. All blood counts are multiplied 109 per litre. Comparisons are either by a Wilcoxon signed-rank test or a chi-squared test as appropriate.
References
Abstract 0040 – E-Poster EPM.016
Clinicians’ attitudes and perceptions to use of ultrasound in percutaneous dilatational tracheostomy placement
1Kings College Hospital
2Guys' and St. Thomas' Hospital Trust
3Heart of England NHS Foundation Trust
Percutaneous dilatational tracheostomy is becoming an increasingly common procedure in the intensive care department. It is also performed by ENT surgeons in the theatre setting. The use of ultrasound assessment of the neck is variably utilised. Some of the proposed benefits for the use of ultrasound include preassessment of anatomical structures, real-time identification of needle position during the procedure and also identifying post procedural complications.1
The aim of this study is to identify the views of clinicians performing percutaneous dilatational tracheostomy as to the perceived value of use of ultrasound in this procedure and subsequently to ascertain whether there is a need for a formal training programme with regard to the use of ultrasound in this context.
A 15-item survey was distributed to trainee research networks and hospital email networks via Survey Monkey. Individuals were asked to respond only if they had any experience with placement of percutaneous dilatational tracheostomy. Opinions were expressed using five-point Likert scale questions and free text comments. The results were gathered over a five-week period.
There were 114 respondents to the survey across the three different specialities (intensive care, anaesthetics, ENT surgery). The majority of respondents identified that ultrasound pre-assessment of the neck in percutaneous dilatational tracheostomy improves safety of the procedure (Figure 1), though there was a clear split between senior and junior clinicians. Real-time assessment was not identified as being necessary or common practice. There was a general feeling that images need not be stored as part of the clinical record or for expert review. Respondents identified that significant barriers to more widespread use of ultrasound were lack of training and lack of experience.
It is evident from clinician’s responses in this survey that ultrasound is perceived to improve the safety of the procedure. There is, however, wide variation in practice, with more experience practitioners less likely to use it. Potentially in the future, a more formal structured training programme may be developed for use of ultrasound in percutaneous dilatational tracheostomy, perhaps as part of wider training and assessment for all procedural ultrasound.
Reference
Abstract 0021 – E-Poster EPM.017
Respiratory physiotherapy and lung ultrasound: A service evaluation of a training programme
Blackpool Teaching Hospitals NHS Foundation Trust
Lung ultrasound (LUS) has the potential to enhance respiratory physiotherapist’s clinical examination and inform treatment. However, there are very limited formal training opportunities for physiotherapists within the United Kingdom. This paper evaluates a training programme initiated on two intensive care units.
Six physiotherapists commenced a four-phase competency-based LUS training programme. The training programme followed the LUS component of the Core Ultrasound Intensive Care accreditation pack offered by the Intensive Care Society. The training included an introductory course, supervised scan sessions, completion of scan report logbook and finally a triggered assessment. The first 100 scan reports to be submitted by the physiotherapy trainees were collected and analysed, and each trainee was involved in a peer discussion about their experiences.
Mentor feedback from the first 100 scan reports included advice about optimal scanning depth, artefact identification, avoiding rib shadows, orientation within the thorax/abdomen and lung sliding. Reported barriers for completing the training programme were a lack of access to diagnostic ultrasound machines. This was due to either being in use by other members of the critical care team or time spent away from the critical care unit. Two of the physiotherapists had managerial responsibilities that took precedence over some aspects of the training. As the training progressed, a list of physiotherapy-specific indications to scans began to emerge. At the time of writing, two of the trainees had completed the triggered assessment and had been deemed competent to practice independently.
Physiotherapists can learn LUS, but there are a number of potential barriers for completing an LUS training programme, which need to be addressed. However, the scan reports included a wide range of pathologies including all of the most common identifiable pathologies seen with LUS. The four trainees, who submitted scan reports, were consistently identifying pathologies and artefacts correctly after 10 scans. It is essential that sufficient time is allocated to complete the training and attendance on an introductory course is advised. Finally, it is essential to be under the supervision of an experienced mentor throughout the training process for regular guidance and feedback. This allowed any feedback points to be integrated into practice earlier on during training to improve image acquisition and interpretation.
Once an LUS introductory course has been completed, sufficient time has been allocated and a suitable mentor has been recruited, physiotherapists are able to become independent and competent in lung ultrasound use on critical care.
References
Abstract 0129 – E-Poster EPM.018
Pre-procedural bedside ultrasound scanning of the anterior neck prior to percutaneous dilatational tracheostomy in ICU. Does it change practice?
Warrington and Halton NHS Trust
Percutaneous dilatational tracheostomy (PDT) is a common procedure in intensive care carried out by intensivists, rather than surgeons. PDT is considered safe,1 but it is associated with significant and potentially life-threatening complications, including bleeding.2 The use of portable ultrasound (US) scanning of the anterior neck prior to PDT is described and is recommended as a method for reducing complications.3 We reviewed how often US changed our practice, such that an open surgical approach was chosen instead of PDT.
Three consultants prospectively collected data from all patients in whom they planned a PDT in our unit. All the patients had US screening of their anterior neck prior to the procedure. When pre-tracheal blood vessels were identified which were felt to contra-indicate PDT, the patient was referred to the ENT team for open surgical tracheostomy (ST).
The three consultants planned PDT in 44 patients (23 male) between January 2015 and June 2016. All these patients underwent a bedside pre-procedural US by a consultant intensivist. Seven (16%) of these patients had blood vessels overlying the trachea at the planned level of PDT, detected by US, which resulted in their receiving open surgical tracheostomy in preference to PDT. All 44 patients successfully underwent tracheostomy without intra-operative complications. Six out of seven ST patients had vessels requiring intra-operative surgical ligation ± division of thyroid isthmus; the remaining ST patient’s isthmus was divided, but had no notably sized vessels at the level of ST. One ST patient had a significant post-op bleed on the second post-op day, necessitating urgent return to theatre.
In this case series, 16% of patients had some form of anatomical variance which changed clinical management such that ST was chosen rather than PDT. ST allows direct visualisation of the anatomy and greater control via surgical ligation and diathermy. We hypothesise that US screening reduces the risk of peri-procedural haemorrhage. However, US increases the rate of ST and non-clinically significant delay in performing a tracheostomy due to surgeon and theatre availability. Just as endoscopy has reduced the incidence of tracheostomy tube misplacement during PDT, we believe pre-procedural bedside US may avoid tracheostomy-related haemorrhage.
An US of an anterior neck vessel. Surgical exposure of anterior neck vessel at ST of the same patient.
References
Abstract 0092 – E-Poster EPM.019
Evaluation of the use of transthoracic echocardiography in intensive care within the oncological population
A Harky and
Barts Heart Centre
References
Abstract 0022 – E-Poster EPM.020
Compliance with pharmacological venous thromboembolism prophylaxis in critical care
A Devlin,
Sunderland Royal Hospital
Venous thromboembolism (VTE) is a common and potentially fatal complication in patients with critical illness.1 Prophylaxis using low-molecular weight heparin (LMWH) has been shown to be effective at preventing VTE and should be mandatory unless contraindicated.2 We aimed to audit compliance with national recommendations for LMWH VTE prophylaxis in a single critical care unit.
A prospective audit was undertaken of all patients admitted to the integrated critical care unit (ICCU) at Sunderland Royal Hospital between 29 September and 31 October 2015. The audit references were the Guidelines for the Provision of Intensive Care Services.2 Data were collected from patient electronic and paper-based records. Compliance was defined as the number of days LMWH was administered or appropriately withheld in the presence of a contraindication, as a proportion of the total patient days. Contraindications to LMWH were prothrombin time >30 s, platelet count <50 × 109/L, therapeutic anticoagulation, major bleeding, cerebral vascular accident with the previous two weeks, spinal surgery with 72 h or heparin induced thrombocytopenia. Data collection was registered and approved by the Trust audit office in accordance with Caldicott principles.
Data were collected on 79 patients admitted over the observation period. The median age was 67 (IQR: 51–76), APACHE II score of 20.5 (SD: 8.0), length of stay on ICCU of three days (IQR: 1–6) and median duration of hospital stay 11 days (IQR: 3–19). Prophylaxis in any form was used on 329 patient-days; 86% of this was using LMWH. A total of 363 out of 411 in-patient days (88%) were compliant with guidelines. Reasons for not giving LMWH included anticoagulation at therapeutic dose (38%), platelet count <50 (37%), major bleeding (9%) and prothrombin time >30 s (8%).
There were differences in compliance between day of admission in ICCU (68%) when compared with subsequent days (88% on day 1 and 92% on day 2).
Overall compliance with recommendations with pharmacological VTE prophylaxis was high but lower than previous reports.3 We observed that compliance on day 0 was considerably lower than compliance rates than any other day. This audit was limited by the inadequate documentation of non-pharmacological VTE prophylaxis. It highlights need for improvement in adherence with VTE prophylaxis recommendation particularly in the first 24 h of ICCU admission. Recommendations are made to improve compliance.
References
Abstract 0005 – E-Poster EPM.021
Using ‘normothermia’ as a target for targeted temperature management (TTM) in post-cardiac arrest patients: Friend or foe?
Nottingham University Hospitals NHS Trust
Targeted temperature management (TTM) following return of spontaneous circulation (ROSC) after cardiac arrest improves neurological outcomes. Recent data suggest this is due to avoidance of pyrexia, as opposed to hypothermia. Following this change in the evidence base, we noted ‘normothermia’ was often stated as the target temperature. We audited TTM post-cardiac arrest, before and after this change to assess its impact.
Consecutive cases admitted to ICU in two three-month period were included; one before and one after the change in practice. This consisted of 18 records in the initial period, and 32 in the second period. Four records (two in each period) were excluded due to a non-cardiac cause of cardiac arrest.
Ten (56%) and 21 (65%) in the first and second periods were male and the mean age was 61.4 and 62.3 years, respectively. For four individuals (2 in each period), the initial rhythm was not recorded. In six (33%) and 12 (38%), the initial rhythm was VT/VF and in the remainder it was PEA/Asystole. In 11 (61%) in the first period and 22 (68%) in the second TTM was indicated, and was used. TTM was not used in one (6%) in the period prior to change in practice and eight (25%) cases following the change, though no contraindication was noted.
Following this change in practice, the median maximum temperature during the cooling period increased from 36.4°C to 37.6°C. After the change in local guidance, 5 of 14 patients exceeded 38°C, compared with 1 of 10 before this change. No significant difference was noted in terms of mortality or in neurological outcomes in the two groups.
Changes in the evidence related to TTM post-cardiac arrest had the unintended effect of reducing its utilisation and an increasing the median maximum temperature of those receiving TTM despite no overt change in local guidelines. This may have been due to the uncertainty of parameters surrounding ‘normothermia’. As a result, specific temperature targets have been reinforced at our centre to improve TTM in this patient group.
Guideline to be implemented to enable effective neuroprotection post OOHCA.
Abstract 0031 – E-Poster EPM.022
Temperature control after cardiac arrest in a district general intensive care unit
R Barber and
Lincoln County Hospital
Our inclusion criteria were:
OOHCA
Primary PCI prior to intensive care admission
GCS <8 and ventilated
Targeted temperature management
Our exclusion criteria were:
Non-cardiac aetiology
Death within 72 h
Core temperature not measured
Our data collected:
Cumulative time above 36.5 within 24 h
Cumulative time above 37.5 within 24 h (danger zone)
Cumulative time above 37.5 from 24 to 72 h
All patients arrived at an acceptable temperature from the cardiac catheter lab.
Despite the request for ‘normothermia’, seven of the nine patients became too warm, exceeding 36.5 within the first 24 h. Time spent above this temperature ranged from 1 to 19 h.
Three patients went on to become pyrexial, surpassing a temperature of 37.5.
Beyond the first 24 h, only one patient was maintained below a temperature of 37.5. Two patients spent more than 24 h above this accepted temperature.
The lack of a clear definition of normothermia.
The lack of a timescale to meet temperature targets.
A non-standardised and multiple device method of temperature control.
A reactive rather than proactive means of controlling temperature.
This has led to the formulation of a post OOHCA cooling guideline to be employed on the intensive care unit, in order to help maximise neurological recovery post arrest.
A further re audit once this guideline has been implemented will take place to assess our compliance with the audit standards.
References
Abstract 0217 – E-Poster EPM.023
Glycaemic control in Harrogate District Hospital (HDH) critical care: A quality improvement project
P Antill, L Craver, R Lathey, S Marsh and
Harrogate District Hospital
Stress-induced hyperglycaemia is a well-known phenomenon frequently seen in critically ill patients. Both hyper and hypo-glycaemia are associated with adverse outcomes including increased morbidity, mortality and length of stay.1
Whilst what value range constitutes normoglycaemia in the critically ill patient remains unclear, current literature suggests that whilst uncontrolled levels are unacceptable, tight control may also increase the risk to patients.2 A number of national and international groups have now suggested parameters for blood glucose (BG) levels, recommending that ideal blood glucose should be kept in the range of 6–10 mmol/L with 4–12 mmol/L being acceptable.3
In view of this, we conducted a retrospective audit of glycaemic control on our intensive care unit (ICU) to ascertain our compliance with BG levels between 4 and 10 mmol/L. Prior to the audit, we had no protocol for glycaemic control, with practice varying widely in terms of acceptable blood glucose levels and time taken to start insulin.
The audit was conducted on all admissions between January and May 2015, encompassing 634 glucose measurements for 42 patients (19 ventilated, 23 non-ventilated). Results demonstrated that 62% of all critical care patients recorded a glucose level out of range (OOR) within the first 48 h with a mean duration OOR of 13.3 h, or up to a mean total duration of 34% of their admission. Of the ventilated patients, 74% were found to be OOR during the first 48 h, compared with 48% of the non-ventilated patients. Five episodes of hypoglycaemia were recorded, of which 80% were in the ventilated group. We identified that patients who spent the first 48 h entirely within target glycaemic range had shorter length of stay than those who spent ANY time OOR (7.1 vs. 11.3 average days, not significant).
In response to this, we have produced an evidence-based document detailing a review of current literature, the reasons for and importance of glycaemic control in the critically ill and a new protocol to attempt to standardize this area of care on our ICU. The document has been produced in conjunction with a multidisciplinary team including pharmacy, diabetologists, diabetic nurses, intensive care nurses and intensivists, to ensure that it can be used by all and is suitable for use. Following the introduction of the document and new protocols, we aim to repeat the audit to assess its effect.
Flowchart for Non-Diabetic Patients
References
Abstract 0148 – E-Poster EPM.024
Audit of targeted temperature management after out of hospital cardiac arrest in a district general hospital intensive care unit
K Bell, S Harris and T Judd
Musgrove Park Hospital
Pyrexia following out of hospital cardiac arrest (OOHCA) is associated with poor neurological outcomes. Previous recommendations following OOHCA should include inducing mild hypothermia.1 New evidence,2 recommends that maintaining an absence of pyrexia is equivocal to creating mild hypothermia in preventing post cardiac arrest neurological damage.
1. Documentation of initial temperature recording.
2. Time taken until there was a documented decision to initiate TTM post return of spontaneous circulation.
3. Documentation of why TTM did not occur, if applicable.
4. If decision was made to implement TTM, were target temperatures achieved?
First, their clinical notes were reviewed and the data entered into a proforma. The second stage consisted of review of all patients’ ICU observation charts.
The graph below illustrates the number of patients whose temperatures were within the target range for post resuscitation care. It also highlighted the percentage of patients who had documented TTM.
The graph below shows the amount of time taken for TTM to start once a patient had a temperature outside the target range.
Previous practice stressed the importance of early targeted temperature management and the use for specific devices, but the change to allowing ‘normothermia’ has allowed patients to develop unchallenged mild pyrexia, outside recommended guidelines.
Target temperature management remains one of the key components of early OOHCA post resuscitation care even though the target had changed.
References
Abstract 0072 – E-Poster EPM.025
2,4 Dinitrophenol: Deadly drug for weight reduction – A fatal case report
S Nallapareddy and
Ipswich Hospital
2,4 Dinitrophenol (DNP) is an industrial chemical, sold over internet as a safe weight reduction pill. This drug reportedly causes rapid weight loss by uncoupling of oxidative phosphorylation and increase in the basal metabolic rate.
DNP has very narrow therapeutic index and high toxicity profile. Hence, medical prescription of DNP was banned in 1938.1 Despite this, it remains widely available, and there has been a significant increase in reported incidents of acute toxicity in the last decade.2
We report a fatal rare case of deliberate acute on chronic overdose of DNP.
A 27-year-old young female presented with recurrent DNP overdose over the preceding one year, consuming similar toxic doses without any serious complications. During this presentation, she was tachycardic but remained hemodynamically stable in the first 17 h of admission. She then suddenly developed malignant hyperthermia (MH) and minutes later, sustained an asystolic cardiac arrest.
The management of cardiac arrest was challenging due to severe muscular rigidity. Intubation and emergency tracheostomy were not possible due to extreme rigidity of tracheal and para-tracheal muscles and resistance to non-depolarising muscle relaxant. Immediately following the administration of dantrolene, muscular rigidity was reversed facilitating endotracheal intubation. There was no response to cardiopulmonary resuscitation and she eventually died.
The case highlights the challenges surrounding the diagnosis and management of acute on chronic dinitrophenol poisoning. Early use of dantrolene is recommended in the management of MH like syndrome with DNP toxicity. There is a possibility that ryanodine receptor gene mutation is linked with MH susceptibility in certain individuals when exposed to DNP. Further research and genetic testing are needed to determine this correlation.
References
Abstract 097 – E-Poster EPM.026
Subcutaneous pentobarbitone overdose
S Billingham1,
1Royal Stoke University Hospital
2The Royal Shrewsbury Hospital
3Princess Royal Hospital
She arrived at the emergency department being bag mask ventilated by the paramedics. She remained neurologically obtunded with a GCS of 6 (E1V1M4) and therefore intubated with a modified rapid sequence induction and transferred to the critical care unit. On examination, she was found to have a 15 cm × 15 cm fluctuating mass on her anterior abdominal wall and this was deemed her likely injection site. It was felt that the large volume of injectate could act as a depot and prolong her sedation; therefore, every effort to expedite drug elimination was sought. After liaising with the national poisons information service (NPIS), she was commenced on multiple-dose nasogastric activated charcoal, and 24 h later, she started haemodialysis. The dialysate fluid tested positive for benzodiazepines. She underwent a CT head and an EEG, which ruled out any intracranial pathology. After a failed extubation and a subsequent ventilator-associated pneumonia, she was successfully extubated 15 days later, though remained confused requiring quetiapine. She was discharged home neurologically intact a month after admission.
References
Abstract 0109 – E-Poster EPM.027
Case presentation: Air there and everywhere!
H Jones,
Morriston Hospital
A 23-year-old male with a history of mental health disease and self-harm presented to the emergency department as a trauma call. His mechanism of injury was unclear, but from minimal history, it was presumed to be a suicide attempt, likely jumping from a cliff.
On initial assessment, he was haemodynamically stable and despite being tachypnoeic had oxygen saturations of 100% on 15 L. Primary survey demonstrated bilateral thoracic tenderness, lumbar spinal tenderness, multiple superficial abrasions and a right frontal haematoma. He also had extensive surgical emphysema extending from the waist to the face.
He was transferred to CT to image his head, neck, chest, abdomen and pelvis. Findings included:
• Gross surgical emphysema
• Large right sided pneumothorax with associated collapse
• Large pneumomediastinum
• Small left apical pneumothorax
• Multiple right-sided rib fractures with a flail segment
• Comminuted right frontal bone and frontal sinus fractures
• Small extradural haematoma
• Air tracking into the thoracic spinal canal at multiple levels: pneumorrhachis
This is an unusual case because amongst the expected and common trauma injuries, there was also a rare finding, which was diagnosed on CT imaging. Pneumorrachis, air contained within the spinal canal1 is a rare and typically self-limiting asymptomatic pathology with multiple aetiologies. However, it is an important diagnosis as it has the potential of spinal cord compression, intracranial and intraspinal hypertension.2 These patients therefore require regular neurological observations and prompt MR imaging if there is any acute deterioration. This is an example of an interesting uncommon pathology, which had not been encountered by the local critical care or spinal teams involved with the case. This patient did not develop any further complications and was discharged from critical care within a few days. Pneumorrhachis is demonstrated very well on axial CT imaging, which once seen will never be forgotten.
CT image demonstrating pneumorrhachis.
References
Abstract 0231 – E-Poster EPM.028
A body to die for
A Holder, R Mahroof and
JVF Critical Care, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust
Throughout the ages, there was increased anxiety about the body image. Dinitrophenol (DNP) is a highly toxic industrial chemical that has been used as an explosive, herbicide and a fat-burning agent. It uncouples the oxidative phosphorylation, disrupts the proton gradient of the mitochondrial membrane and blocks the production of ATP. The energy is lost as heat. It is easily available for purchase from internet and is popular amongst bodybuilders and people trying to lose weight. There are 62 lethal cases reported at present related to DNP use.1 We are presenting a case of successful management of a patient with DNP poisoning.
A 24-year-old male patient was admitted to the Intensive care unit (ICU) after ingestion of 1380 mg DNP. Initially, he was mildly agitated, tachycardic and had temperature of 37.2°C that self-corrected. Twelve hours post-ingestion, the patient developed fever 40.2°C and increased agitation requiring intubation and mechanical ventilation. Hypermetabolism expressed itself with increasing CO2 production that required a minute ventilation of over 25 L per minute, hyperthermia and rhabdomyolysis.
Continuous veno-venous haemodiafiltration was instigated at 20 ml/kg/min. Cooling techniques were applied including a cooling blanket, cooled IV fluids, dialysate and replacement fluids, bladder irrigation with cold 0.9% sodium chloride and ice packs.
Dantrolene was administered at increment doses to 10 mg/kg in total. Hydrocortisone 200 mg, 1 g vitamin C, Glucose 20% and VRIII were added to the therapy.
Phosphate-creatine supplement was commenced as an alternative fuel.
Temperature control was achieved within 8 h. The hypermetabolism persisted over the following two days, at which point the patient was extubated. Then, the glucose-insulin infusion, phospho-creatine supplement and CVVHDF were all ceased.
At day 5 of admission, the patient developed compartment syndrome of his left calf and underwent an anterior and lateral compartment fasciotomy. The patient was discharged from ICU on day 9 since admission and on day 20 went home.
There is no known antidote against DNP. Rapid recognition of the poisoning, admission to ITU and immediate intervention are the key points to successful management.
Reference
Abstract 0256 – E-Poster EPM.029
Prisoner in my own body – Severe case of Guillain-Barré
1Royal Gwent Hospital, Aneurin Bevan Health Board
2Morriston Hospital
3Aneurin Bevan UHB
A 33-year-old male patient presented with acute respiratory failure following a cough and shortness of breath. On presentation, he was in type II respiratory failure with a lobar consolidation. Due to worsening hypoxia and drowsiness, he was intubated uneventfully on ICU.
After 6 h of ventilation, he had significantly improved and underwent a sedation hold from which he failed to wake. Following this, the patient had a CT head which was normal. The history was revisited with the family and it was noted that the patient had been having difficulty walking for the preceding 48 h due to poor coordination, later believed to be significant ataxia. Despite this, the patient had no discernible neurological deficit preintubation and had been combative with normal strength. A full neurological examination revealed a widespread flacid areflexia in all myotomes. During the examination, it was noted that the patients left eyebrow had sustained twitching. It became immediately apparent that these twitches were intentional and that the patient was awake and conscious with profound weakness. An examination of his cranial nerves revealed a complex ophthalmoplegia and nystagmus, but with normal pupillary responses. An LP revealed an isolated raised protein count which leads to a provisional diagnosis of Guillain-Barré. This was confirmed on EMG and nerve condition studies. Anti-GQ1B antibodies were found to be positive, consistent with the Miller Fischer variant of Guillain-Barré.
The patient was treated with IV immunoglobulin. In total, the patient received five days of 0.4 gkg/day of Octogam, which equated to 35 g per day. His clinical course was complicated with multiple ventilator-associated pneumonias; however at six months, he was liberated from ventilatory support and discharged for rehabilitation.
Guillain-Barré is a progressive peripheral polyneuropathy, with onset of symptoms normally over days to weeks. It classically presents with a symmetrical ascending motor paralysis with or without sensory and autonomic features1 and is usually preceded by a triggering event, predominantly an infection1. In contrast, this case involved a rapid onset of severe weakness over a number of hours, atypically of Guillain-Barré.
This case represented a diagnostic challenge given the speed and severity of onset. Additionally, it provided a huge challenge from a communication perspective for the whole MDT dealing with a locked in patient. Despite the poor prognostic factors, the patient made a surprising recovery and continues to improve in a local neurological rehabilitation centre.
Reference
Abstract 0219 – E-Poster EPM.030
Early fluid resuscitation in major burns: Change in practice over a decade
1Chelsea and Westminster Hospital
2Imperial College London
The aim of this study was to examine the changes in fluid prescribing over a 10-year period in a regional burns centre. It is well established that major burns patients require early resuscitation with large fluid volumes. Current practice is to base this on the use of the Parkland Formula. Excess fluid over time (fluid creep) is considered to be harmful in these patients; however, optimal resuscitative volumes have yet to be determined. Furthermore, controversy still persists over the role of colloids in the initial resuscitation phase.
We performed a retrospective review of patients admitted to our centre with a >15% total body surface area (TBSA) burn over three one-year time periods (2005/2010/2015). Data were collected on patient demographics, fluid input and output and type of fluid administered over 24 h from admission. The study was performed in accordance with local ethical/governance standards.
There were no significant demographic differences between the groups, but there was a trend towards lower TBSA% burn since 2005 together with lower admission lactate and base deficit (Table 1). Overall, there is no significant difference in the amount of fluid administered to patients, as expressed in ml/kg/TBSA% (Table 2). There was a trend towards a reduction in use of colloid suspensions. There was one abdominal emergency in the 2015 group.
Groups.
Fluid data.
References
Abstract 0185 – E-Poster EPM.031
Is there still a role for a two-tier trauma response within a trauma unit? A service evaluation of caseload and outcomes of trauma resuscitation room cases at a large District General Hospital
Treliske, Royal Cornwall Hospital Trust
Triaging hospital trauma response represents a significant challenge. Overcalling trauma may stretch the system, while undercalling risks missed injuries, delayed treatment and potentially substantial harm. Recent NICE guidelines suggest trauma units should operate a single-tier trauma call system.1 Our hospital operates a two-tier system with either an emergency department (ED) trauma team including a consultant or middle grade with a senior house officer, or a multi-disciplinary hospital trauma team. This audit reviews the role of this two-tier system within our trauma caseload.
This was a retrospective audit, reviewing trauma patients admitted to the resuscitation room in the ED of a large District General Hospital during July 2016. Patient notes were reviewed for pre-alert, initial allocation of trauma tier, injuries and outcome, and appropriateness of trauma call based on presentation in ED as per the Hospital Trauma Team Activation Criteria.
In total, 24 patients were included. Median age was 36 (range: 5–91). All patients presented with blunt injury.
In total, 7 out of 24 (29%) trauma activations did not follow network criteria. All three patients who had no trauma call should have received an ED trauma call. Three out of 7 (43%) patients receiving an ED trauma call should have received a hospital trauma call. All hospital trauma calls were appropriate, though 4 out of 14 were based on team leader discretion.
Planned trauma response.
Actual trauma response.
One patient admitted to ICU, following a hospital trauma call, died later, giving a total trauma case mortality of 17%, and a hospital trauma call mortality of 28.5%. Total home discharge rate was six from to 24 (25%).
Trauma caseload is difficult to measure in a busy ED. Overall, there is a trend towards undercalling trauma despite the presence of appropriate triage criteria. The role of the ED trauma call remains debated given mixed outcomes between home discharges and ICU admissions. These numbers are missed on current trauma registry data but measuring them is crucial to running and improving a functioning and efficient Trauma Unit within a network.
Reference
Abstract 0259 – E-Poster EPM.032
Respiratory effects of rib fixation in patients with severe chest trauma
1Royal London Hospital
2Barts Health NHS Trust
Abstract 0006 – E-Poster EPM.033
Mobilising the immobile: Emergency cervical spine management in the obtunded trauma patient
Queen Elizabeth University Hospital
Abstract 0080 – E-Poster EPM.034
Small volume fluid infusion with balanced or un-balanced fluids: Does it make a difference?
D Benoit, M Crivits, M Hemeryck, E Hoste,
University Hospital Ghent
Abstract 0264 – E-Poster EPM.035
The use of patient/family diaries in the cancer critical care unit: A prospective mixed methods study
S Dolan1, K Gull2, C Lucas2, G O'Gara2 and N Pattison2
1Kings Health Partners
2The Royal Marsden
Survivors of critical illness often take time to recovery physically and psychologically from their critical care experience.1 There is increasing evidence suggesting that the use of a patient- or family-led diary with entries by nurses, doctors and allied health professionals may help the patient ‘fill in the gaps’ by making sense of a time they have forgotten, or have confused or frightening memories of, and can aid the patient and their family members in the recovery after critical illness.2,3
References
Abstract 0228 – E-Poster EPM.036
Is volume-based feeding the way forward for UK critical care patients?
1London Northwest Healthcare NHS Trust
2City, University of London
References
Abstract 0122 – E-Poster EPM.037
An investigation into current UK practice for screening adult patients with a tracheostomy tube for dysphagia
1Royal Free London NHS Foundation Trust
2Kingston University and St George's University of London
Patients with tracheostomy tubes are at risk of aspiration and swallowing problems (dysphagia) and because of their medical acuity, complications in this patient population can be severe.1 It is well recognised that swallow screening in stroke significantly reduces potential complications by allowing early identification and appropriate management of patients at risk (by health professionals), thereby reducing delays in commencing oral intake and preventing unnecessary, costly interventions by speech and language therapists (SLTs).2
However, there is no standardised swallow screening for the tracheostomised population and there is a paucity of literature regarding either current or best practice in this area.
The aim of this study was therefore to investigate current United Kingdom (UK) practice for swallow (dysphagia) screening for adult patients with tracheostomy tubes and to explore and describe health professionals’ perceptions of their current practice/current systems used.
A mixed methods approach was adopted, comprising a semi-structured online questionnaire and recorded follow-up telephone interviews. Participants were SLTs, nurses and physiotherapists working with patients with tracheostomies. Responses were analysed to determine current practice with regard to swallow screening. Thematic analysis of interviews allowed further exploration and clarification of the questionnaire findings.
Two-hundred and twenty-one questionnaires were completed. Approximately half (45%) of the participants worked in trusts with formal swallow screens, whilst the remainder used a variety of other approaches to identify patients at risk, often relying on informal links with multidisciplinary teams (MDT). In line with current evidence, patients with neurological diagnoses and a tracheostomy were consistently referred directly to speech and language therapy. Only a quarter of questionnaire participants thought their current system was effective at identifying patients at risk of swallowing problems.
Eleven questionnaire participants were interviewed. They highlighted the important role of MDT team working here, emphasising both its strengths and weaknesses when working with these patients.
Current practice in the UK for screening patients with a tracheostomy for swallow problems is varied and often sub-optimal Despite the evidence base for enhancing outcomes, MDT working is still perceived as problematic. A swallow screening tool for use with this population, to enhance MDT working and ensuring that practice fits in line with current evidence, may improve patient safety and care.
References
Abstract 0263 – E-Poster EPM.038
An exploration of delirium incidence, CAM-ICU compliance and patient characteristics in Cancer Critical Care Patients
Royal Marsden NHS Foundation Trust
Delirium is characterised by disturbed consciousness, cognitive function or perception with acute onset and fluctuating course and is associated with poor outcomes despite how it can often be prevented and treated.1 Critical care patients may experience delirium due to organic issues (i.e. hypoxia, sepsis, mechanical ventilation) or non-organic (i.e. lack of sleep, noise, communication difficulties, pain). It is the most common neuropsychiatric complication in patients with cancer related to disease, opioid use and side-effects of treatment.2
Describe the proportion of cancer ICU patients who experience delirium. Explore any potential predictive factors for delirium in critically ill cancer patients, such as: cancer treatment, performance status, admitting problem, gender, age, LOS, hypoxia.
References
Abstract 0196 – E-Poster EPM.039
Implementing a restrictive blood sampling strategy in a large tertiary cardiac ICU
1Papworth Hospital NHS Foundation Trust
2Queen Elizabeth Hospital
Anaemia and the requirement for blood transfusion after cardiac surgery have been shown to negatively impact on morbidity and mortality.1 Cumulative diagnostic blood sampling can be a significant cause of haemoglobin attrition, particularly in patients in intensive care. We sought to determine whether implementation of a low volume restrictive blood sampling strategy in a large cardiothoracic intensive care unit (CICU) reduced the burden of sampling loss and whether this was associated with higher haemoglobin on discharge from CICU or reduced need for red blood cell (RBC) transfusion.
In March 2014, standard adult 3 ml arterial blood gas (ABG) syringes were replaced with 1 ml paediatric ABG syringes in CICU. We then retrospectively evaluated ABG sample volume, discharge haemoglobin and RBC transfusion in 8887 patients admitted to CICU before and after the intervention. Patients were categorised by length of stay (LOS) as short (<72 h), medium (3–7 days), long (7–30 days) or prolonged (>30 days), and timing of transfusion as early (<72 h), mid (3–7 days) and late (>7 days).
Overall daily blood loss due to ABG sampling was reduced from (mean ± SD) 67±3.7 ml in 2013 to 40±9.7 ml and 39±8.8 ml in 2014 and 2015, respectively (p < 0.0001). This effect was observed with equivalent significance in all LOS categories. In long-stay patients, this equated to a mean reduction in ABG-related blood loss of 280 ml and 1160 ml for those with LOS >30 days. There was a reduction in total RBC transfusion during the study period, 4637 units in 2013 to 3222 units and 3296 units in 2014 and 2015 (p < 0.0003). The greatest number and greatest reduction in RBC transfusions were within 72 h of critical care admission (p < 0.0001). Differences in mid and late transfusion did not reach statistical significance. Reduced ABG sampling did not correlate with reduced transfusion (p = 0.3) or with haemoglobin concentration at ICU discharge in any LOS group (p = 0.24).
Implementation of a low volume restrictive sampling strategy was associated with a significant reduction in ABG sample volume but not in discharge haemoglobin or need for RBC transfusion in patients admitted to CICU. Reasons for this may include changes in transfusion trigger following publication of the TITRe2 study in 20152 and continued blood sampling for other diagnostic tests. We plan to re-evaluate after introduction of further measures to reduce burden of diagnostic sampling. Maximal benefits may be seen when strategies to reduce blood sampling volume are used as part of a multimodal programme of patient blood management.
References
Abstract 0052 – E-Poster EPM.040
Introduction of a point-of-care ultrasound teaching programme for novices with skill retention on follow-up
A Chadwick, P Duggleby, B Griffiths, J Hughes,
John Radcliffe Hospital
Point-of-care ultrasound (PoCUS) is a valuable skill that allows for easy detection of pathology and therapeutic intervention at the bedside. Research has demonstrated that not only can it be incorporated into diagnostic algorithms to make them more accurate and improve patient management1 but that the skills required can be picked up by novices relatively easily.2 It has already been incorporated in the curricula of several North American medical schools.3
With this in mind, we designed and implemented a new half-day ultrasound course for Oxford University medical students and foundation year doctors. It was intended for those with little prior knowledge or experience of ultrasound practice and designed to meet the following learning objectives:
To understand the physics underpinning ultrasound Recognise the uses, advantages and disadvantages of PoCUS in the acute care setting Identify vascular structures on ultrasound and differentiate arteries from veins Safely use ultrasound to assist in difficult peripheral venous cannulation
The course consisted of a set of lectures followed by practical skill stations. The lectures focused on ultrasound physics, basic echocardiography, thoracic ultrasound and abdominal ultrasound. The practical skill stations delivered hands-on teaching in ultrasound-guided vascular access and basic chest and abdominal ultrasound.
We used pre-course and post-course questionnaires, to assess improvement in attendees' knowledge of ultrasound physics and uses as well as improvement in confidence to use ultrasound for peripheral venous cannulation. We then conducted an online follow-up survey at an average of four months post course to assess retention of knowledge and skills.
A total of 48 candidates (37 medical students and 11 junior doctors) attended the course. There was a significant improvement in understanding of ultrasound physics (53.2% to 97.8%, p < 0.0001), uses of PoCUS in acute care (78.7% to 100%, p = 0.0011) and confidence in using ultrasound to gain peripheral venous access (8.5% to 85%, p < 0.0001). At follow-up survey an average of four months later, 20% had used ultrasound in clinical practice since attending the course. Understanding of ultrasound physics and use in the acute setting was maintained (100% for both) and confidence in using ultrasound for peripheral cannulation was reasonably well maintained at 66.7% (See Figure 1).
The course demonstrated that teaching basic PoCUS to medical students and FY doctors is highly valued by candidates and results in greater understanding of the principles of ultrasound and confidence in using it for basic applications. Self-reported knowledge and skills are retained at four-month follow-up.
References
Abstract 0053 – E-Poster EPM.041
Point-of-care ultrasound training on ICU: An analysis of accreditation programmes worldwide
1Hospital General Universitario de Castellon
2John Radcliffe Hospital
Although championed by enthusiasts, the use of ultrasound in ICM (CCUS) has lagged behind that of other specialties including emergency medicine. Two international expert statements acknowledged the challenges of obtaining appropriate training in CCUS and aimed to describe the components of competence with specific goals of training and skill development. The lack of a uniform formal training structure and programme is a recurring issue across Europe and worldwide, posing the questions of whether scans have been appropriately performed/reported, and whether there exists proper clinical governance to ensure a high standard of care.
Comparisons were made between the accreditations programmes available worldwide which included those from Americas, Europe and UK. Analysis was made as to which ultrasound modalities were included in the programme as well as the process of training/accreditation within the programme.
A total of seven accreditation programmes were included for analysis. Common themes included the need for didactic teaching, direct supervision and maintenance of a logbook. There were significant difference in modalities, competencies, number of minimum scans and assessments between the programmes.
There is considerable variation in currently available programmes which may lead to confusion. Crucially, even within modalities, there is not an agreed method as to how best to train and acquire the skills of CCUS. It is crucial that appropriate protocols are formalised to ensure robust clinical governance. We await a Europewide CCUS accreditation by ESICM as part of its CoBaTRICE (Competency in ICM training) initiative. There is a need for a unified and consistent CCUS programme for ICM. National organisations such as the ICS should lead and engage such an initiative.
Abstract 0094 – E-Poster EPM.042
Focused intensive care echocardiography (FICE): A retrospective study in quality assurance to guide future practice on ICU
1University of Cambridge
2Papworth Hospital NHS Foundation Trust
Abstract 0084 – E-Poster EPM.043
HOCUS POCUS: Hands-on cloud uploading service for point-of-care ultrasound
South Warwickshire NHS Foundation Trust
The use of ultrasound imaging in acute settings such as the intensive care unit or emergency department is becoming increasingly common. Point-of-care ultrasound (POCUS) has the advantage of rapidly providing relevant information at the bedside without the risks of patient transfer or radiation.
For these reasons, a number of ultrasound accreditation schemes have been developed to encourage and facilitate the training of clinicians in these skills.
The focused intensive care echocardiography (FICE) accreditation scheme provides a recognised structure of training, reporting and accreditation in echocardiography. The core ultrasound skills in intensive care (CUSIC) accreditation has also been introduced and provides competency assessments for lung, abdominal and vascular ultrasound.
FICE and CUSIC accreditation requires that a trainee’s initial scans are directly supervised by their mentor until basic competencies are met. Subsequent scans can be undertaken without direct supervision, but these must be stored for review with the mentor. Barriers to this process can include the absence of a trained mentor at a trainee’s location, lack of shared time slots during the working day and trainees rotating to different sites whilst undergoing accreditation. These logistical issues can make accreditation challenging to achieve in the specified timeframe.
For this reason, we developed a cloud-based ultrasound logbook which allows the trainee to upload anonymised cases along with comments and questions to their mentor. The mentor is then able to view the trainee’s images remotely, scrutinise the trainee’s comments and respond to any questions. Discussions over video-conferencing software can also take place, whilst ultrasound clips are being viewed simultaneously by trainee and mentor. This provides the possibility of ‘face-to-face’ education without geographic constraints. It also provides an easy route for mentors to seek second opinions or encourage shared learning amongst their peers.
Attention must be given to security issues throughout this process and all patient-identifiable data must be removed from all files prior to use. This project was undertaken with the input of our Trust’s Caldicott guardian ensuring confidentiality requirements were adhered to.
We believe that the use of secure cloud software and video-conferencing technology can improve access to ultrasound training opportunities and improve trainee and mentor experience. The platform and user interface will continue to be improved and could soon be available to be trialled on a larger scale.
Abstract 0069 – E-Poster EPM.044
The staff focused always event
J McCann and
Warrington and Halton NHS Foundation Trust
The Always Event® is a concept that was initially developed in the US by the Picker Institute. They are events in a patient’s experience that are deemed to be so important to patients and their family members, that we, as heathcare providers, must aim to perform constantly for every patient every time.2 The patient or the relative determines the event.
In Warrington ICU, we decided to extend this concept and have a staff focused always event. Instead of asking patients and relatives, we could ask staff members for ideas for always events to help improve patient safety. We followed the existing Always Event® structure that Events should be important, evidence based, measurable and affordable.
We closed the audit loop and saw an 18% improvement in patients being nursed head up. All cases of VAP on our ICU are recorded. Following the introduction of the always event in May 2016, we have not had any further cases of VAP.
References
Abstract 0110 – E-Poster EPM.045
Methods of Central venous catheter securement and chlorhexidine dressing use: A survey of practice across Southern England
Royal Brompton Hospital
Given the ubiquity of central venous catheter (CVC) use in the critical care setting, significant uncertainties remain regarding how best to minimise the incidence of central line associated bloodstream infections (CLABSIs) and localised infection. A bundle of simple interventions was demonstrated to be of benefit1 and has been widely adopted. A recent Cochrane Library systematic review2 suggests that sutureless securement devices and chlorhexidine impregnated dressings reduce the risk of CLABSIs, but it is unclear how widely this recent research has changed practice.
We conducted a survey of 87 adult critical care units within six Critical Care Networks across southern England. We spoke to a senior nurse or doctor at each unit via telephone (August–September 2016), asking them how CVCs inserted in their unit are typically secured, and whether a chlorhexidine-impregnated dressing or patch (e.g. Biopatch®) is routinely used for the insertion site.
We received complete responses from all 87 units contacted.
Ninety-one per cent (79/87) use sutures to secure CVCs, and 9% use sutureless securement devices. The majority of those using a sutureless securement device use the Statlock® brand (7/8). No units use tape/adhesive strips as their method of securement.
Forty-five per cent (39/87) use chlorhexidine-impregnated dressings or patches routinely at the insertion site. Of the units using localised chlorhexidine devices, one uses them only for femoral insertions sites and another uses them only in immunocompromised patients. All the other units (37/39) use them in all patients and for all insertion sites.
These findings demonstrate that practice within the majority of units surveyed is divergent from the advice of the most recent systemic review on this topic. This may be due to the relatively low quality of the evidence favouring sutureless securement devices and chlorhexidine-impregnated dressings – a point that the Cochrane team themselves also allude to. Our findings suggest that further research into these interventions is warranted.
References
Abstract 0165 – E-Poster EPM.046
Audit of transfusion triggers in intensive care unit at Southmead Hospital Bristol
1Royal Devon & Exeter Hospital
2Southmead Hospital
Patients often receive blood transfusions in the critical care setting. There is increasing evidence of harm as well as benefit. Various studies such as the 1999 Transfusion Requirements in Critical Care (TRICC) study, 2002 Anemia and Blood Transfusion in critically ill patients (ABC) study have shown that a restrictive approach (transfusing at <70 gdl) is at least as effective and possibly superior to liberal approach (transfusing at <100 gdl). A higher transfusion trigger may be beneficial in patients with ischaemic stroke, traumatic brain injury with cerebral ischaemia, acute coronary syndrome (ACS) or in the early stages of severe sepsis. Other recommendations include transfusion of one unit of blood at a time, with assessment of symptoms after transfusion and post transfusion Hb levels.
Southmead hospital intensive care unit is a 40-bed unit. An audit conducted in September 2015 on the unit showed that a significant proportion of patients were transfused at inappropriate triggers. A sample of n = 28 units of transfusion were identified in September through data collecting forms and 14.29% patients received inappropriate transfusion triggers. Where more than one unit were used, only 25% transfusions showed evidence of post transfusion Hb checks.
Issues identified in the first audit were highlighted to the department. This included lack of clear accessibly standardised department trust guideline and lack of awareness amongst junior staff. Changes were implemented including the introduction of a new online trust guideline for blood transfusion in critical care, teaching and training sessions for junior doctors at department induction. Repeat cycle conducted in March 2016 identified n = 35 units of transfusion. There was a reduction in inappropriate triggers to 11.43%. Post Hb checks during transfusion where more than one unit were used, there was an improved rate of 40% (Table 1).
Comparison of audit results from September 15 to March 2016.
References
Abstract 0176 – E-Poster EPM.047
Documentation of FASTHUG – Improving the standard of care
Addenbrooke's Hospital
Reference
Abstract 0240 – E-Poster EPM.048
Is it worth instigating excellence reporting in a district general hospital?
E Rizal and
Heart of England Foundation Trust
A short survey was devised and distributed amongst multidisciplinary staff in an ICU. The questions related to staff morale, attitudes towards incident reporting and whether or not they felt rewarded for doing good work. The survey allowed for free comments to support user ratings.
A total of 37 responses were collected over a three-week period from seven different health professional titles including consultants, junior doctors, SAS doctors, nurse practitioners, nurses, allied health professionals. The results are described in graphical form below. Generally, morale was low, highlighting issues with short staffing and lack of autonomy. When asked about whether they felt rewarded for doing a good job, the response was mostly equivocal. Some professionals felt valued and others felt unsupported or unappreciated. Staff were also in equipoise about IR1 forms involving themselves. Those who commented, however, felt supported but distressed. IR1 was described as being a disciplinary tool to seek blame. Within the department, it was felt that learning occurred from negative experiences rather than positive ones with comments including the lack of positive feedback and misbelief that system failures have been rectified by protocols.
The uptake of survey responses is limited in that nursing staff numbers are potentially under-represented. Nonetheless, the general consensus of responses was sufficient to allow us to close the survey at the three-week mark. The department is in agreement about having low morale amongst some of the health professionals. IR1s are generally viewed negatively and it is felt that little learning occurs from this.
Comments have described an interest in positive feedback suggesting that a formal tool for recognising great work can encourage positive learning and morale within the department. A pilot of IR2 – Excellence reporting is warranted within the department to ascertain if the staff experience can be improved. This is currently ongoing and the results are looking promising.
Abstract 0227 – E-Poster EPM.049
Detection of occult post-extubation laryngeal injuries during routine fibreoptic endoscopic evaluation of swallowing
B McGrath, S Wallace and
University Hospital of South Manchester
The consequences of trans-laryngeal intubation range from glottic oedema to permanent laryngeal injury. Abnormal vocal cord mobility may contribute to post-extubation dysphagia, aspiration, poor cough and ineffective vocalisation. Laryngeal injury and dysphagia often remain undetected in the critically ill and the impact on aspiration, secretion clearance and oral feeding is not well understood. Whilst the use of fibreoptic endoscopic evaluation of swallowing (FEES) by speech and language therapists (SLT) is important for early detection and management of dysphagia and aspiration, FEES can also detect occult laryngeal pathology. The aim of this study was to investigate the frequency and nature of laryngeal injuries detected by routine FEES in post-extubation critically ill patients, referred to SLT with suspected dysphagia.
All post-extubation ICU patients with suspected dysphagia were referred to SLT and underwent a clinical bedside swallowing assessment. Standard FEES indications included signs of aspiration, excess laryngeal secretions, neurological or neuromyopathy symptoms, slow respiratory wean or dysphonia. Laryngeal injuries were described and aspiration rated using the penetration aspiration scale. Patient data including intubation duration and tracheostomy status were also recorded at the time of FEES, with time taken to achieve normal oral dietary intake recorded post-FEES as part of routine SLT follow-up. Mann Whitney U test was used to compare the differences between groups.
FEES by SLT is important for detecting occult, unsuspected laryngeal injuries which may contribute to dysphagia, dysphonia, secretion clearance, aspiration and impact on dysphagia rehabilitation, feeding and airway management decisions.
References
Abstract 0026 – E-Poster EPM.050
Albumin creatinine ratio predicts medical intensive care unit outcome of patients with hepatic encephalopathy
1Internal Medicine, Al-Azhar Faculty of Medicine, Al-Azhar University
2Clinical Pathology, Al-Azhar Faculty of Medicine, Al-Azhar University
Abstract 0158 – E-Poster EPM.051
Epidemiology of pulmonary embolus in England 2015/2016
N Bunker
Barts Health
More patients are being diagnosed with pulmonary embolus (PE) as the availability and resolution of CT pulmonary angiograms has increased but the incidence of fatal PE has remained static.1 Using hospital episode statistics (HES), the epidemiology of PE can be accurately described and may allow treatments to be more effectively targeted.2
HES were interrogated from April 2015 to April 2016 including all patient episodes in England. All patients with a primary or secondary diagnosis of PE were identified and then further divided according to age, gender and whether they were an emergency patient or an elective patient. The ICD-10 and OPCS codes were analysed for these patients to describe the characteristics of patients who developed a PE and identify the commonest surgical procedures leading to PE.
There were in total 20,603,622 hospital episodes from the financial year 2015/2016, of these 95,631 were coded with primary or secondary diagnosis of PE (0.46% of all HES data). This is an annual incidence of approximately 1.5 per 1000. Of these, 9417 patients were elective patients presenting for major surgery whereas the remainder were emergency admissions; 51.75% of PEs were in females with the majority of the imbalance in the emergency cohort.
PE was the primary diagnosis in just over half the episodes and the remainder were divided between common medical diagnoses such as stroke (1165 episodes), heart failure (1070 episodes), COPD (1419 episodes), pneumonia (5420 episodes) and malignancy (5500 episodes). On the basis of recorded OPCS codes for those patients presenting with PE, the commonest surgical procedures that resulted in PE were: total knee replacement, total hip replacement, hysterectomy and endoscopic removal of bladder lesion. Inferior vena cava filters were only inserted 239 times.
A total of 5400 patients with PE died or 5.65% of all those presenting with PE, although it is not possible from these data to identify whether they died directly due to the PE or another diagnosis.
PE continues to be a common diagnosis and represents significant morbidity and mortality. The majority of patients present as emergencies with metastatic cancer a major risk factor. Orthopaedic and pelvic surgeries continue to be the commonest cause of acquired pulmonary embolism after elective surgery. Further analysis of the HES dataset may help further elucidate the high-risk groups and allow consideration of more complex, invasive preventative strategies.
References
Abstract 0088 – E-Poster EPM.053
Early deaths in the ITU: The importance of patient selection and senior decision-maker input
D Bose and G Reid
Warwick Hospital
Intensive care is a high-cost resource for which demand will likely always exceed supply. In our seven-bedded mixed high-dependency/intensive care unit, we felt that a significant number of patients died within 24 h of admission and that the decision to admit them may have been inappropriate. We sought to analyse the factors leading to their admission in order to better inform our decision-making in best utilising the limited resource of intensive care.
A retrospective analysis was undertaken of the medical notes, ITU charts and ICNARC data for all patients (n = 26) who died within 24 h of admission in a one-year period (January–December 2014). The patient demographics, reason for admission, grade of referring doctor, whether the admitting consultant was informed and whether s/he was an intensivist, day and time of referral, time to admission from referral, time of death, APACHE II score, organ systems supported and cause of death were all recorded in a spreadsheet. Not all data were available for every patient.
Patients were predominately elderly (mean age: 70.4, range: 49–90) and 57% were female. Almost half (43%, n = 6) of patients died within 12 h of admission. Mean APACHE II score was 32.3 (range: 16–40), with mean APACHE II predicted mortality of 66.98% (range: 34.5%–97.3%). The majority of patients (77%, n = 20) were admitted outside 8 am to 5 pm on weekdays without onsite consultant presence, with 60% (n = 12) of these at weekends. The referring doctor was identified as a consultant in 50% of cases. In only 26.9% of cases was the patient was discussed with a consultant intensivist prior to admission.
In conclusion, a significant number (6.7% of total annual admissions) of elderly patients with poor predicted mortality are admitted out-of-hours without discussion with a consultant intensivist. Early senior decision-making may improve patient selection, ensuring that the limited resource of intensive care is used appropriately for the patients likely to benefit the most.
Abstract 0198 – E-Poster EPM.054
Medical emergency teams’ hospital outcomes in a day (METHOD) – An international study of rapid response teams shows impact of frailty on outcomes of deteriorating patients
C Subbe1,
1Ysbyty Gwynedd
2Austin Hospital
3Stanford University
4North Tyneside General Hospital
5University College London Hospitals
Data collection was obtained prospectively from 1134 patient records and dates of admission ranged from April 2014 to March 2016. Data collection and classification of outcomes have been described before.1
Physiological abnormalities at the time of rapid response review were described by the following parameters (mean/standard deviation); a respiratory rate of 24 (9), oxygen saturations of 93 (9)%, systolic blood pressure of 122 (34) mmHg, heart rate of 101 (28) bpm; 398 patients had an altered mental status and 869 patients received supplemental oxygen. The median degree of physiological abnormality translated into a level of the National Early Warning Score2 of 8.
The age range was from 16 to 101 years. Median age was 71 years and 289 patients were more than 80 years old. Thirty-nine per cent of patients were classified as frail according to the Clinical Frailty Scale (Figure 1).
At the time of 30-day follow-up of patients with known outcomes, 321 (28%) patients had died, 514 (45%) had been discharged to their own home, 75 (7%) to care homes, 150 (13%) patients were still in hospital, and 60 (5%) patients had been transferred to rehabilitation or other hospitals.
Impact of frailty on clinical outcomes.
Clinical Frailty Scale.
References
Abstract 0239 – E-Poster EPM.055
After critical care outreach and VitalPac – What next for reducing post-ICU mortality?
P Falkous1, R Howarth1 and D Monkhouse2
1Northern School of Anaesthesia
2James Cook University Hospital
Abstract 0046 – E-Poster EPM.056
The impact of recent changes in the number of critical care beds on patients’ mortality – A preliminary study
London North West Healthcare NHS Trust
The National Health Service (NHS) across the United Kingdom faces a lot of challenges. The recent changes in the healthcare system have resulted in a decreased number of critical care beds and an increased workload and duties in one of the hospitals in our trust. In Northwick Park Hospital (NPH), there has been a reduction in five high-dependency unit (HDU) beds in the first quarter of 2016. The aim of our work was to assess the impact of reduction of critical beds on mortality of patients and on the number of interhospital transfers.
The primary outcome was to assess the trends in mortality of patients admitted to critical care settings in NPH before and after the change in the number of critical care beds. Secondary outcome measures included the number of interhospital transfers of the acutely ill and the severity of patients admitted to the critical care unit. Data were collected in the first quarter of 2015 and 2016, respectively, and compared to those from Ealing Hospital (EH) where there was no change in critical care provision. The first set of results was collected before the change resulting in reduction of HDU beds. They were used as a reference point and were compared to the second set of data collected after the change.
In NPH, the first quarter mortality rate was higher in 2016 in comparison to 2015 (27.22% vs. 18.24%, p = 0.0565). A similar trend was observed in the number of transfers of critically ill patients. It has increased from 30 to 41 (p = 0.2284). The third evaluated effect of the decreased number of HDU beds was the severity of those admitted to the ITU which showed an increased APACHE II score (20.19 vs. 23.59, p = 0.0010). All the studied parameters remained relatively unchanged in the EH.
The impact of changes in critical care provision on patients’ mortality is well described in the current literature.1 Interhospital transfer of critically ill patients is also associated with an increased mortality.2,,3 The results of our study have shown an increase of mortality and number of transfers in NPH. Furthermore, more severe ill patients required ITU admission which could be a result of delayed admissions due to higher requirements for critical care. The above-mentioned results might be the effect of a reduced number of critical care beds as similar trends were not observed in a hospital from the same area but no changes in HDU bed availability.
References
Abstract 0077 – E-Poster EPM.057
Outcomes of haematology patients admitted to intensive care
Heart of England NHS Foundation Trust
Historically, there has been a reluctance to admit patients with haematological malignancy to intensive care units (ICU) with studies citing poor outcomes.1 More recent studies have shown improved outcomes; a 2009 study utilising ICNARC (Intensive Care National Audit and Research Centre) data from 178 ICUs found an ICU mortality of 43.1% and a hospital mortality of 59.2%.2 Critical illness amongst these patients is often due to complications of treatment with current guidance recommending onsite ICU depending on treatment offered.
Birmingham Heartlands Hospital is a large district general hospital with a level 3 haematology facility which has an active autologous and allogenic stem cell transplant programme. In 2015–2016, there were 1443 ICU admissions in total with average length of stay of two days in level 2 patients and four days in level 3 patients who were emergency admissions. The aim of this audit was to identify mortality and admission demographics in haematology patients.
Forty-four patients were identified as being admitted with a haematological diagnosis to our ICU between August 2014 and August 2016 using data from ICNARC coding and electronic records. The average age was 57 years, and of those admitted, 47.7% (n = 21) survived to ICU discharge. In-hospital mortality was 66%. Amongst those who survived to ICU discharge, 71.4% were alive at hospital discharge and 66.7% were alive at 30 days post discharge. Survival at six months was 25% amongst all patients.
The average length of admission was 6.3 days (range: 0.3–39.7 days), and amongst survivors to ICU discharge was 8.01 days. Median length of hospital admission prior to ICU admission was 13.5 days. Median APACHE II score was 26 (range: 8–37); median score for non-survivors was 29. Tewnty-three patients required level 3 care with 26% surviving to 30 days post-discharge. 54.5% of patients required one organ support, 25% two organ support and 20.5% three organ support.
This has demonstrated higher mortality amongst haematology patients with increased length of stay compared to unit average. A 2012 study at a UK specialist cancer intensive care unit showed mortality of 38.2% and hospital mortality of 51.4%.3 Our audit showed worse ICU and hospital mortality with a higher APACHE II score and longer admission prior to ICU admission. This combination could explain worse outcomes amongst our cohort and may suggest earlier admission would be beneficial. We are currently extending our dataset to five years and plan to analyse for further prognostic variables.
References
Abstract 0024 – E-Poster EPM.058
A retrospective analysis of haematology patients admitted to ICU at Maidstone Hospital
J Cassell
University Hospitals Coventry & Warwickshire
Advances in the treatment of haematological malignancies have resulted in increased survival for these patients, but also potential complications which may result in ICU admissions. The study aimed to identify the pattern of use by haematology patients in a general 14-bedded ICU in a DGH.
Data collection was concurrent and retrospective using hospital intranet systems from December 2014 to May 2015.
In total, there were 24 ICU admissions for 21 patients, mean age was 65 and 62% were male. For the 21 patients, there were at least 11 different haematological diagnoses, of which 42.8% were lymphoma. Mean length of ICU stay was 7.5 days. Ten of the admissions were for respiratory reasons and nine for renal replacement therapy. ICU mortality was 21.7%, with an overall in-hospital mortality of 42.1%. Patients comprising 60.8% were discharged to another ward within the hospital; 20.8% patients were ventilated and 50% had renal replacement therapy.
The overall mortality rates in our DGH are similar to those published for other haematology patients. The published in-ICU mortality rates for haematology patients are higher than for a general ICU population. Length of ICU stay for our haematology patients at 7.5 days was also longer than for general patients at 2.3 days. Discharge to another hospital ward was lower than for general ICU patients (83%). Fewer patients were ventilated in the Maidstone population than in published literature and appreciably more had renal replacement therapy.
Abstract 0260 – E-Poster EPM.059
Long-term survival outcomes in acute kidney injury patients receiving renal replacement therapy on intensive care in a UK teaching hospital: A comparison of ethnicity data
1Sheffield Teaching Hospitals NHS Foundation Trust
2Whiston Hospital
Census data reports wide variations in health between ethnic groups. Health outcome inequalities have been shown for example in Pakistani women, whereas other ethnic groups have lower rates versus the White British group.1
Acute kidney injury (AKI) affects 40% of critically ill patients, with UK data reporting 5% needing renal replacement therapy (RRT).2
We evaluated survival outcomes in AKI patients receiving RRT on our ICU in relation to ethnicity demographic data.
Data were collected on all AKI patients receiving RRT on our ICU, between 2008 and 2013. This included demographics, ethnicity coding and patient survival at ICU discharge, in addition to outcome data at 28 and 90 days and 12 months for each ethnic group. There were 17 different ethnicity codes.
Ethnicity codes.
Two hundred and ten White British patients had survival data recorded; 117 (55.71%) survived to 28 days, 94 (47.24%) to 90 days and 87 (43.94%) to 12 months. Of the patients in the ‘other’ group with survival data 71.4% survived to 28 days, 71.4% to 90 days and 71.4% to 12 months. Overall for those with ethnicity coding, 414 (66.7%) patients survived to ICU discharge, with 368 (59.3%), 341 (55%) and 308 (49.6%) patients being alive at 28 and 90 days and 12 months, respectively. For the entire cohort of 620 patients, overall patient survival at end of follow-up was 43% with 64.6% having renal recovery at one year.
Our results showed that the largest ethnic group receiving RRT was White British, but that these patients had worse outcomes than those in minority ethnic groups. There also appears to be a higher proportion of females in the ‘other’ group versus a male preponderance in the White British group. These differences may be related to the relatively small sample.
References
Abstract 0216 – E-Poster EPM.060
Use of the MAP: Catecholamine ratio as a de-resuscitation trigger
1West Suffolk Hospital
2Addenbrooke's Hospital
The concept of de-resuscitation is of growing interest given the near-ubiquity of positive fluid balance and oedema in survivors of septic shock and the association with increased mortality and morbidity this brings. It is suggested that active removal of excess fluid once the disease process is resolving may improve outcomes. Malbrain et al. recently helped popularise the ‘ROSE’ model, with phases of septic shock split into resuscitation, optimisation, stabilisation, and evacuation.1 A negative fluid balance is targeted in the evacuation (or even stabilisation) phase; this process includes ‘late goal directed fluid removal’ where negative fluid balance is actively sought via diuretic therapy or even haemofiltration. Whilst these authors reserve the term ‘de-resuscitation’ for LGFR in the ‘E’ ±‘S’ phases of their model, colloquially the term is more loosely used in reference to the point at which active fluid removal is deemed possible and desirable.
However, de-resuscitation is conceptually defined; it remains a practical uncertainty when it should commence. Malbrain et al. suggest triggers including ↓P:F ratio, ↑EVLW, ↑IAP, positive cumulative fluid balance and/or total body weight. No threshold values are offered. Undoubtedly, these parameters should be considered but the absence of reliable and repeatable criteria make the triggering of de-resuscitation wholly subjective. The human factor problems in the delivery of critical care are well known; these include poor communication, poor handover, inconsistent plans being formulated, etc.2 The current reality is that objective de-resuscitation is aspirational, but the absence of it risks patient harm.
We recently introduced the concept of MAP:vasopressor (M:V) ratio.3 This metric combines cardiovascular (CV) data into a single descriptor which may help describe the outcome of CV interventions (e.g. fluid challenge/offload). We have subsequently included M:V into routine display via our clinical information system. Anecdotal use has led to interest in M:V as providing an objective metric in the triggering of de-resuscitation. An example is provided (Figure 1), in this case using MAP:catecholamine ratio. In this case, the trough M:V was 130 mmHg/µg/kg/min representing the most critical point of CV compromise. The subsequent rising M:V equates to CV recovery; we anecdotally speculate that an M:V of 500 mmHg/µg/kg/min may be a suitable de-resuscitation trigger. Alternatives include (for example) M:V improving to 3× the trough value, a threshold gradient of M:V rise, or a fixed time period from M:V trough. We advocate the evaluation of M:V as a putative de-resuscitation trigger via further research.
References
Abstract 0017 – E-Poster EPM.061
Redesigning the ICU portable emergency bag
M Chahal and
Royal United Hospital
Abstract 0250 – E-Poster EPM.062
Central line length and patient height – Testing a simple matrix
Dumfries and Galloway Royal Infirmary
Central venous catheterisation (CVC) remains a routine procedure in the critical care setting.1 As such, it is important that these catheters lie in the correct position in a major vessel to prevent avoidable complications.2,3 Ideally, line tips would lie in the lower third of the superior vena cava; however, this can prove difficult to judge on chest radiograph alone with location based on the patient’s own anatomy.4
To improve line placement, we created a line length guide based on insertion site and height of the patient using two height ranges for simplicity.5 To test this, we performed a prospective audit of central line insertions over a three-month span comparing length of line inserted and final position on chest radiograph. Ideal position was judged by defining a ‘Green Zone’ of one posterior rib height above the carina to two vertebral body units below the carina.4 The final position of the central line tip was then measured on chest X-ray in comparison to this and length either long or short was noted. Our practice is to insert the line to the hub and avoid using the secondary fixation device; however, the length of line protruding out of the skin was audited and taken into account for corrected line length.
Out of the 71 total lines audited, after corrections, 46 (64.8%) lines adhered to our created ‘ideal length table’; of these lines, 40 (87.0%) were optimally positioned within the green zone compared to six lines outside this zone. The matching lines which were too short were more likely high RIJ insertions (3/4 lines) with the remaining short line being a 12.5 cm low RIJ which was only 1 cm short likely due to patient body habitus. Lines which were too long were one low RIJ insertion and one medial subclavian insertion.
Lines which match matrix, % correctly positioned in zone.
References
Abstract 0252 – E-Poster EPM.063
Health records on ICU – Is digital really the future? A comparison of three systems in different critical care areas in a single trust
Oxford University Hospitals NHS Foundation Trust
The sheer volume of data on the ICU is overwhelming and arguably beyond the processing power of the average human. Electronic clinical information systems (CIS) have been advocated as the way forward, in order to appropriately manage but also analyse the volume of information available. However, the implementation of CIS alongside electronic patient records (EPR) across ICU is not universal and is subject to local variation. Reasons for non-adoption of technology are also varied.
As part of a quality improvement project, we compared three critical care areas in the same trust which utilise different ways of recording health records (one paper-based vs. two electronic platforms). A variety of matrix was used to measure the quality of the records. A neutral party was used to adjudicate legibility. The results of this work would help in the decision on whether or not to move towards EPR.
Our findings showed that EPR were superior to traditional paper records in a number of key areas, e.g. legibility, date and time recording and identifying consultant care (see Table). Those areas where they were not superior were on account of intrinsic software limitations which could be improved with better templating of data entry.
Reasons for non-implementation of CIS included cost, disruption of work flow (especially for visiting teams) and lack of sufficient IT support. Distraction of clinician’s time away from clinical duties has also been cited.
Our audit showed that from a governance prospective, EPR is superior to paper records. However, the decision to move towards EPR/CIS needs to be individualised to the particular ICU taking into account work flow and logistics. Other advantages of EPR/CIS such as built-in clinical decision support modules, improve auditing ability and notification prompts should also be considered.
Abstract 0235 – E-Poster EPM.064
The innovation of an online patient pathway for those requiring extracorporeal membrane oxygenation (ECMO) for severe acute respiratory failure (SARF)
Royal Brompton and Harefield NHS Foundation Trust
ECMO for SARF is currently centrally funded in five centres within England.1 This means patients, and their families, referred for ECMO travel across large distances to receive this specialist treatment. Therefore, this requires a high level of coordination to refer and transfer a patient in a timely and efficient way. The current referral system is paper which requires a significant amount of administrative time and also may cause patient information to be lost.
The ‘Hospital to Home’ ECMO patient pathway innovation aims to address an identified gap in service provision. The pathway was built out of the success of the children’s national LTV pathway, which has been adapted for the ECMO population as they are of a similar small and complex cohort of patients. It follows the patient through their whole journey from referral for ECMO, repatriation and subsequent rehabilitation until discharge home, crossing organisational boundaries. This will join up the communication processes, allow the sharing of information on a single platform and avoid the duplication of information between professionals. It is designed to use a multidisciplinary approach and enables the capture of patient level data that can be used for service improvement, audit and resource management. It has been designed to be agile which allows detailed levels of viewing permission and interactions with the pathway.
The foreseeable outcomes of the pathway include timely and efficient transfer of information that should enable prompt decision making, ensuring that patients receive the most appropriate treatment quickly. This has the potential to reduce each patient’s hospital stay and therefore reduce cost. The plan to assess these outcomes will be to use analysis of length of stay, number of patients requiring ECMO, staff and patient satisfaction, and audits.
Reference
Abstract 0171 – E-Poster EPM.065
Outcomes after in-hospital cardiac arrest in an LMIC hospital with a nurse-led rescue team and availability of parameters for early warning scores
P Athapattu1, A Beane2, N De Silva3, P De Silva1, A Dondorp4,
1National Intensive Care Surveillance, Ministry of Health
2Network for Improving Critical Care Systems and Training
3Monaragala District General Hospital
4Mahidol Oxford Research Unit
5Department of Clinical Medicine, University of Colombo
In-hospital cardiac arrests result in mortality of over 60% even in high income settings where dedicated resuscitation teams are functional. In low- and middle-income countries (LMICs), cardiac arrest outcomes are less well known, with limited evidence pointing to poorer outcomes. This study describes the characteristics of cardiopulmonary resuscitation practices in a District General Hospital (DGH) in an LMIC where a cardiac arrest nurse responder had been deployed and the availability of physiological variables needed to calculate selected early warning scores.
Availability of physiological variables and relevant EWS-MEWS1, NEWS2, SEWS3, HOTEL4, PARS5, LEWS6, ASSIST7, PEDS8, CART9 ViEWS10 qSOFA.11
Movement and outcomes of patients after resuscitation.
A total of 173 patients were reported to the cardiac arrest team during the study period, of whom 151 were 18 years or older and were thus included in the analysis. Most cardiac arrests to which the cardiac arrest nurse was summoned happened during the day shifts (7 am–7 pm) and 45 (30.4%) were in the weekend (Friday 7 pm to Monday 7 am).
Overall, out of the 150 patients who had CPR during the study period, 52 (35.1 %) patients were discharged alive from the hospital. A total of 74 (47.2%) patients had return of spontaneous circulation (ROSC) after CPR. Out of them, only 31 were admitted to ICU and others remained in the original clinical area. Survival at hospital discharge was similarly not statistically significant between those treated outside an ICU after ROSC and those who were treated in an ICU, though a higher proportion of those treated in ICUs died.
As a conclusion, the limited availability of simple physiological parameters makes validation and deployment of EWS for early detection of deteriorating patients difficult and reinforces the need for acute care skills training for healthcare teams in LMIC settings. EWS and rapid response system in this LMIC may need to be setting adapted.
References
Abstract 0183 – E-Poster EPM.066
A retrospective, propensity-matched, cohort analysis of patients retrieved to the Royal Adelaide intensive care unit
M Finnis1, S Gluck2,
1Discipline of Acute Care Medicine, The University of Adelaide
2SAAS MedSTAR, 16 James Schofield Drive
3Faculty of Health Sciences, The University of Adelaide
South Australia has a widely distributed population which raises significant challenges to the provision of healthcare, particularly to those who are critically ill or injured. Since March 2009, the State retrieval service has been provided by SAAS MedSTAR, which undertakes approximately 5500 retrievals per year, often from remote and austere locations. We speculated that lead in delay and limited rural healthcare resources would lead to increased mortality in retrieved patients admitted to a metro ICU when compared to a non-retrieved cohort.
Demographic data by study cohort.
Data are analysed as: number (%) – Chi-squared, median (25th percentile, 75th percentile) – Wilcoxon rank-sum, mean (95% confidence intervals) – Student’s t-test.
APACHE III – Acute Physiology and Chronic Health Evaluation – Version 3j as per the Australian and New Zealand Intensive Care Society, Adult Patient Database.
Odds-ratios and 95% confidence intervals (CI) for univariate logistic regression for retrieved patients against hospital mortality.
Probability OR = 1 by Wald statistic.
APACHE III – Acute Physiology and Chronic Health Evaluation – Version 3j as per the Australian and New Zealand Intensive Care Society, Adult Patient Database.
Population numbers from Australian Bureau of Statistics Census data.
Data only analysed for non-metropolitan retrieval locations.
During the study period, 1597 (14%) of 11,641 index ICU admissions were retrieved by MedSTAR. Demographic details are listed in Table 1.
Of those retrieved, 295 (19%) were primary, from-scene retrievals and 1074 (67%) were rural or from interstate with a median (IQR) radial distance 226 (97,315) km. The intervals from retrieval activation to arrival on scene and ICU admission were 0.9 (0.5, 1.5) h and 5.4 (3.5, 8.3) h, respectively; the time on location was 0.8 (0.5, 1.1) h.
Univariate factors associated with hospital mortality are listed in Table 2.
The average ‘treatment’ effect for retrieval using propensity score matching was OR (95%CI), 1.04 (1.011, 1.062), p = 0.005.
Our analysis revealed a small but significant negative treatment effect associated with retrieval; the cause of this effect is unclear. Our multivariate analysis shows rural location and time from activation to arrival at scene to not be contributing factors. The significant, paradoxical, difference in mortality from team activation to arrival in ICU remains unexplained.
Retrieved patients were shown to make-up approximately 15% of the workload of the RAH ICU and retrieval was associated with a small (4%) but significant relative increase in hospital mortality. Future studies designed to identify mortality benefits for different retrieval interventions would require very large study numbers or would have to target higher mortality subgroups.
Abstract 0180 – E-Poster EPM.067
The ‘weekend effect’ in critical care – An ubiquitous trans-Atlantic problem?
J Blundell1, F James2, A Herbert2, S Laha2 and N Plummer2
1King's Lynn NHS Foundation Trust
2Lancashire Teaching Hospitals NHS Foundation Trust
The ‘weekend effect’, excess morbidity and mortality in patients admitted to hospital at weekends and out-of-hours,1 has recently been used to drive NHS reform despite debate regarding the cause of this apparent disparity. Critical care should be somewhat sheltered from this; however, international studies have shown a wide variation in outcomes. Healthcare in the US is often regarded as an ideal system; hence, we aimed to assess whether the ‘weekend effect’ existed in US critical care environments.
Weekend and out-of-hours (outside 7 am–7 pm) admissions were more likely to be an emergency case, although severity scores were only marginally different. Relative risk of in-ICU mortality was 1.31 (p < 0.001) for weekend and 1.35 (p < 0.001) for out-of-hours admissions, with persistent excess mortality at one year.
Figure 1 shows the mortality rates at weekend and out-of-hours for patients admitted as emergencies or electively. Emergency patients showed no clinically significant difference in severity scores, and both weekend and out-of-hours admissions were associated with a small increased risk of in-ICU mortality (RR: 1.11, p = 0.009 and RR: 2.00, p < 0.001 respectively), which persisted to one year. However, the difference between weekend and out-of-hours admissions was more marked for non-emergency patients. Both groups were more severely unwell when assessed using APSIII and had longer ICU and hospital stays. Weekend and out-of-hours admissions demonstrated an excess in-ICU mortality (RR: 3.84, p < 0.001 and RR2.54, p < 0.001, respectively), persisting to one year.
References
Abstract 0161 – E-Poster EPM.068
Time burden of critical care outreach referrals
The Royal Wolverhampton NHS Trust
Reason for referral and mean referral time burden.
Note: Values are in minutes (95% CI).
References
Abstract 0107 – E-Poster EPM 069
Does introducing a rapid response team reduce the frequency of avoidable cardiac arrests?
1Western Sussex Hospitals NHS Foundation Trust
2Cambridge University Hospitals NHS Foundation Trust
The notes of all individual patients with confirmed cardiac arrests outside a critical care area between February and July 2015 were examined (n = 55). Factors assessed included: timing of medical review, timing of observations, MEWS scoring pre-arrest, involvement of outreach and a global subjective assessment of management. Comparisons were made with 151 historical control cardiac arrests between May 2010 and January 2013, with analysis for statistical significance by chi square (SPSS version 24).
1. The cardiac arrest rates are shown below. The observed changes were not statistically significant.
2. Most patients (32 of 55) received appropriate pre-arrest care. All were reviewed by a doctor within the preceding 24 h. The last medical review had in most cases (32 of 55) been carried out by a consultant or SpR. Thirty-eight of 55 had observations recorded within the preceding 4 h. Categorisation of the management of these patients is shown below. Insufficient data existed to compare these outcomes with historical controls.
3. In the post-RRT group, 81% of arrested patients had an MEWS score of 4 or less preceding their arrest. 64% of historical controls had an MEWS score of 4 or less and this difference was significant at the p < 0.05 level (two-tailed). This result is confounded by poor recording of observations in the control group.
Abstract 0062 – E-Poster EPM.070
“Deep experience is never peaceful” (Henry James): Raising awareness of noise sources in the intensive care unit
1University of Oxford
2Oxford University Hospitals NHS Foundation Trust
The intensive care unit (ICU) is a cacophony of unfamiliar noises that can be terrifying for patients. Up to 75% of patients experience ICU-acquired delirium.1 These patients remain longer in hospital and can have long-term psycho-social problems after they go home.2
Noise levels in ICUs are around 60 dBA. This is about as loud as conversation in a busy restaurant. Sudden peak sounds can be >120 dBA. The World Health Organisation recommends that noise levels in patient care areas should be <35 dBA with peaks <40 dBA.
Using the AEBCD framework,3 patients and staff planned an intervention to address noise levels in the ICU. This included replacing metal bins, initiating an alarms management policy to encourage individualised patient settings, and raising awareness through new teaching materials developed with clinical educationalists.
An e-learning module includes patient interviews and evidence-based information on the effects of noise. A short experiential teaching session includes sounds and activities common in the ICU. This gives staff the opportunity to experience the ICU from the patient perspective.
To date, 68 staff members have completed the training; 100% found the experience useful, 96% said they would change their practice after the session and 93% thought staff in the ICU could do more to create a less stressful environment for patients. Self-identified changes include increased patient reassurances, reducing unnecessary patient interactions and promoting a quieter environment. From a word list generated from patient interviews, the most frequently selected words (>25 each) by participants after the teaching session are ‘worrying’, ‘frightening’, ‘stressful’, ‘confusing’, ‘uncomfortable’ and ‘alone’.
Feedback demonstrates that staff are unaware of how much noise normal activities create. We are now evaluating the effectiveness of the training as well as its popularity.
References
Abstract 0018 – E-Poster EPM.072
A quality improvement intervention to reduce noise on the ICU
Intensive Care Unit, Queen Alexandra Hospital
There is increasing evidence that interrupted sleep on the intensive care unit is linked to the development of delirium which, in turn is an independent risk factor for morbidity and mortality. One of the causes of interrupted sleep is excess noise.
We undertook a pilot study measuring noise levels every second at a variety of central and patient-side locations over a two-week period.
Our results broadly agree with other data published on this topic and show that our noise levels are well above the levels recommended by the World Health Organisation of 35 dB.
Our unit is quieter at night than in the day with a trough time period of 11 pm to 5 am at which time our average noise levels are approximately 52 dB. During the day, average noise levels are approximately 57 dB. There is no difference in noise levels measured at the patient-side versus those measured at the charge station.
Our peak noise levels are also in excess of desired levels with peaks above 70 dB ranging from 100 to 1750 times per hour, and levels above 80 dB occurring between 5 and 48 times per hour.
As a result of these findings, we have appointed four nurse ‘noise champions’ which, in conjunction with the procurement of some visual noise activated warning signs will help to raise awareness of excess noise levels. We have also established a collaboration with the University of Portsmouth to explore the development of technology that will provide a personalised patient environment to reduce noise levels and potentially utilise white noise.
Abstract 0251 – E-Poster EPM 073
Critical noise levels: Still an issue for patient care?
Heart of England NHS Foundation Trust
References
Abstract 0173 – E-Poster EPM.074
Use of a multi-faceted teaching programme to improve the confidence of ICM doctors in bronchoscopy
1London North West Healthcare NHS Trust
2Chelsea and Westminster Hospital NHS Foundation Trust
3Imperial College School of Medicine
Participants from all disciplines significantly increased their confidence, with pre- and post-course medians of 5 and 8, respectively (p < 0.01). This effect was more pronounced amongst junior (CT1-ST4) than senior participants (ST5 and above), reporting confidence score increases of 4 compared with 3 (p = 0.03).
Limitations include the small sample size, confidence ratings as a subjective measure of skill and course participants as the study population. Further work is important to correlate confidence and competence and investigate the long-term efficacy of multifaceted training. Nonetheless, an opportunity exists to improve training in ICM bronchoscopy.
References
Abstract 0123 – E-Poster EPM.075
Building a fully controllable ECMO moulage into a high-fidelity simulation lab
Harefield Hospital
Abstract 0128 – E-Poster EPM.076
Simulated Tricky Trips Transfer Course (STricT) – Using simulation to develop an interprofessional multi format transfer course
1Ashford & St Peters Hospital
2Royal Surrey County Hospital
There are an estimated 11,000 transfers that occur throughout the UK per year1 which result in a three-day increase in length of stay for critically unwell patients.2 Due to the centralization of services and increase of tertiary centers, these transfers are becoming more common.
There are well-defined guidelines for transfer by the AAGBI and the ICS. With the development of a specialized simulation ambulance, an opportunity arose to develop an interprofessional simulation-based transfer course.
Candidates are assigned to interprofessional groups involving doctors, ODPs and nurses before undergoing simulated exercises where technical and human factors skills are explored.
The simulated scenarios include a packaging station where the team ready a patient for a time critical head injury transfer. They then prepare a level 3 ICU patient and transfer them to CT experiencing a variety of clinical and logistical issues on the way. An inter-hospital transfer is replicated by taking a patient on an immersive simulation ambulance drive.
Each session concludes with a faculty-facilitated debrief session focusing on technical and human factors.
The course has a HEKSS funding bid for development and has formal endorsement from the ICS.
We will use candidate feedback to continue to improve the course and hope to expand our scenario and MCQ bank.
Abstract 0197 – E-Poster EPM.077
Safe critically ill transfer training (SCITT) – ‘networking’ for training
1Cheshire & Mersey Critical Care Network
2Greater Manchester Critical Care Network
3Lancashire & South Cumbria Critical Care Network
4East Lancashire Hospitals NHS Trust
Effective training for the transfer of critically ill patients has historically provided several challenges for critical care networks and NHS Trusts:
How can performance across regions be standardised? How can the training needs of different critical care health professionals be addressed? How can training and revalidation for the extensive number of staff working with critically ill patients be facilitated? How can training that addresses the needs, clinical environment and equipment issues of individual trusts be provided?
It remains an often expensive and challenging dilemma for critical care networks when attempting to meet ICS Standards regarding safe transfer of critically ill patients.
The Safe Critically Ill Transfer Training (SCITT) Programme is a collaborative endeavour between three adjacent Critical Care Networks to provide standardised transfer training for multidisciplinary professionals in critical care. The number of doctors and nurses requiring training in the region is approximately 185 and 2329, respectively. This novel approach of integrating an e-learning platform, locally delivered simulation training and logbook evaluation provides a viable solution for provision of training in an efficient and affordable manner. Use of a regional ‘Steering Group’ allows the training to remain relevant and current while fostering vital collaborative working between Networks with common practices and goals.
It is suggested that this model may be easily reproducible to provide a national training package while allowing bespoke adaptations that incorporate regional differences in practice and locally identified learning from critical incidents. It has potential to address the transfer training needs of other professional groups such as paramedics, and both emergency medicine & acute medicine practitioners. This standardised, locally delivered training with regionally controlled governance presents a model conducive to the extensive training needs of a modern healthcare service.
SCITT is an example of inter-Network collaboration promoting partnership in the delivery of patient safety, clinical effectiveness and patient experience in line with the Department of Health’s Five Year Forward View.
Abstract 0199 – E-Poster EPM.078
Integrating regular multidisciplinary ‘in-situ’ simulation into the education programme of a critical care unit. How we do it
Manchester Royal Infirmary, Central Manchester Foundation Trust
Simulation-based teaching has become a major form of effective education.1 In addition to practical skills and stress inoculation, ‘in-situ’ simulation further introduces multidisciplinary educational aspects of communication skills, team-working and human factors. It also allows teams to test their efficiency in a controlled manner in real time and in the real location, therefore showing any system errors that could be improved.
Learning from inter-departmental experience,2 we describe the development of a multidisciplinary ‘in-situ’ simulation education programme in an inner-city tertiary critical care unit. We discuss lessons learnt, human resources required, tailoring of scenarios to local needs, identification and prevention of errors in patient care. We also outline future plans for a fully integrated multidisciplinary and multi-modal education programme.
We describe our experiences of various barriers in the implementation of this programme. In particular, we have demonstrated how the allocation of dedicated time to nurses and doctors for facilitating the whole simulation process has enabled regular delivery of effective sessions despite a busy critical care. Institutional barriers to new forms of learning were challenged, through the increased frequency of simulation sessions and now it is accepted as part of the weekly education schedule.
With a dedicated multidisciplinary steering team, multidisciplinary ‘in-situ’ simulation can become an invaluable resource to improve patient care, stress response, teamwork and learner satisfaction.
We believed that having a multidisciplinary team consisting of nurses and doctors helped the simulation team transition and develop into a weekly occurrence and gain more involvement from all areas of the unit. The multidisciplinary team allowed more insight on specific training needs of nurses and doctors together and apart. Improved communication and teamwork were the main areas identified as outcomes following the simulations, and it allowed the team to work more efficiently together in stressful situations. Team members were able to improve on non-technical factors such as closed-loop communication, effective delegation and assigning roles.
We outline the overwhelmingly positive feedback for simulation sessions. We also describe the use of tailored feedback via certificates and e-mail to participants.
References
Abstract 0085 – E-Poster EPM.079
Stress at the front door: Impact on critical care service provision
Mid Essex Hospitals NHS Trust
The demand for A&E services has increased from 1.5 million monthly attendances in 20101 to over 2 million in 2016.2 Acute medical admissions are increasing the burden on already scarce resources with recurrent breaches in key performance targets since 2014.3
While it is clear that the pattern of acute medical admissions is changing, the nature of these changes has not been quantified. The effects of increasing acute medical admission on critical care workloads have also not been studied.
We performed a retrospective, demographic analysis of all acute medical admissions from 2005 to 2015 in Broomfield Hospital, a district general hospital with tertiary specialist services. Data were collected for a randomly-chosen two-week period, every year over the study period. Demographics were also collected for all medical admissions to Broomfield Intensive Care Unit from 2009 to 2015.
Simple descriptive statistics were used to summarise the data. Adult admissions were classified based on age using predefined intervals. A two-sample test for equality of proportions was used to compare proportions with continuity correction as appropriate. Beta regression was used to model and predict future admission proportions.
Acute medical admissions have increased exponentially since 2010, mainly due to increased A&E admissions. The number of primary care referrals has remained static.
The increase in admissions are almost entirely due to patients aged 75 and above (Figure 1). The proportion of acute medical admissions with age more than 75 has increased from less than 13.6% in 2005 to 50.5% in 2015. This proportion is predicted to increase to 71.8% (95% CI: 65.0–78.6) by 2020, if current trends continue (Figure 2).
There is a concomitant increase in the number of patients admitted to the intensive care unit. However, this is only associated with a modest change in the mean age at admission (1.73 years, 95% CI: 0.03–3.41, p = 0.04).
There has been a sustained increase in acute medical admissions from 2010, mainly due to increased elderly care admissions. There has been a concomitant increase in admissions to ICU, but without an increased proportion of elderly admissions. The increase in acute medical admissions could have resulted in offloading of activity to ICU in all age groups and not just the elderly. The projected increase in elderly admissions in the future has significant implications for ICU resourcing.
Change in proportions of acute medical admissions across different age groups. There is a sustained increase in patients aged 75 and above. While the proportion of patient in 55–75 age group is decreasing, the absolute numbers are increasing. The predicted increase in elderly admissions (75 years and above) as a proportion of total admissions over the next five years.
References
Abstract 0213 – E-Poster EPM.080
Trust-wide standardisation of supplementary equipment for the intra-hospital transfer of critically ill adult patients
Queens Medical Centre, Nottingham University NHS Trust
References
Abstract 0262 – E-Poster EPM.081
Emergency endotracheal intubations in the critically ill: An analysis of a UK trainee database
Hammersmith Hospital, Imperial College
In-hospital emergency intubations are performed in patients with limited physiological reserve, with limited planning, and in occasionally in suboptimal locations. Intensive care doctors are faced with increasingly complex patients with deranged baseline physiology and complex conditions who are disproportionately likely to experience airway difficulty, presenting challenges to airway safety in ICU.
We undertook a prospective, observational study of endotracheal intubation performed by critical care trainees in United Kingdom to identify practice and complications. An online logbook, ‘MedELogbook’ (www.medelogbook.com) designed for ICU junior doctors to log their referrals, transfers, procedures and resuscitation calls was created in 2013 Although designed primarily as an aide for training, it provides a novel approach to obtaining an insight into the working patterns of ICU junior doctors.
In total, 663 intubations were reported in the past two years by over 100 trainees in the UK. The incidence of clinically significant complications associated with emergency intubations is not insignificant. The overall complication rate of 10%, with desaturation being the commonest complication. Complications were associated with intubation grade rather than ASA grade.
The number of emergency intubations only dropped significantly during weekday out of hours (20:00–07:59). This may represent adequate planning and anticipation of the deteriorating of patients requiring emergency intubation. Emergency/outreach staffing levels on weekends is crucial given the similar frequency of emergency intubations on weekday daytime and weekends. Approximately two-thirds of intubations were carried out in the intensive care unit.
Despite ward patients being older, with disproportionately higher ASA grades and with the lowest GCS prior to induction (i.e. high risk), almost all cases where capnography were not used were on the wards. This is of concern following NAP4 and ICS guidelines.
Thiopentone and suxamethonium are not the induction agents and NMB most commonly used for emergency intubations. Instead, propofol (often in combination with other sedatives) and rocuronium are the commonest induction agents and NMB used, respectively. This may represent better familiarity with propofol. Our data suggest that rocuronium now the most popular choice of NMB among intensivists being used in 54% of our intubations, with suxamethonium accounting for 31%. A change in the drug choices for emergency intubations may intuitively be safer, though clinical trials are needed to confirm this.
Abstract 0265 – E-Poster EPM.082
Repatriation process for level 2/3 patients from tertiary cardiothoracic intensive care unit to local DGHs
The Essex Cardiothoracic Centre
1. To assess our current practices in relation to repatriation
2. To identify reasons for delays and its impact on service provided
3. To make recommendations for future improvements/interhospital collaborations
1. Data collected retrospectively from 2013 to 2016 (July)
2. 75 patients identified
3. All patient details in relation to repatriation noted
4. Data analysed
5. Results will be presented on the Centre’s Audit Meeting
1. Out of 75 patients, there were 56 male patients and 18 female patients
2. Thirty-nine cardiology admissions and 36 cardiothoracic surgical patients
3. Median length of stay for cardiology patients until the transfer was 17 days
4. Median length of stay for cardiothoracic surgical patients until the transfer was 32 days
References
Abstract 0106 – E-Poster EPM.083
Improving written handover from intensive care to general wards, through the introduction of a standardized discharge documentation form
1East Kent Hospitals University
2King's College Hospital
ITU resources are limited, and prompt patient discharge is necessary to ensure good patient flow. Robust handover processes are essential to prevent errors, adverse events and re-admissions. Poor handover may negatively affect morale of ward doctors. Studies have shown that standardized forms are effective in improving handover documentation.1 In 2007, NICE issued specific guidelines on the minimum datasets to be included in ICU discharge documentation.2
The lead author aimed to address concerns about variable quality and availability of ITU discharge documentation, highlighted by foundation doctors in local feedback sessions.
An initial survey of 17 wards showed that 64% were satisfied or highly satisfied with ITU discharge documentation, 79% felt confident or highly confident to rely on the information provided and 50% found that discharge documentation very or extremely helpful when formulating management plans.
Baseline audit of 36 discharge summaries revealed that 86% of patients had a ‘free text’ discharge summary. Although a summary of ITU admission was well documented, compliance with vital information, such as medications, nutrition and infection status, discussions with patient/relatives and verbal handover was poor or non-existent.
A standardized discharge proforma was introduced. Follow-up audit of 44 discharge summaries showed that, with the new proforma, 97% of discharge summaries had a medication list, 84% a nutrition plan/status, 45% relevant infection considerations, 55% included a summary of discussions with patient/family and 71% documentation about verbal handover. Ten months later, this new proforma remained in use, and re-audit of 15 discharge showed similar results, suggesting sustained improvement was achieved.
Despite this, follow-up surveys showed that overall satisfaction rates remained unchanged, suggesting that additional factors might play a role.
This project highlights how standardised proformas can be an effective intervention to improve compliance with ITU discharge documentation. The project, which was presented at the International Forum on Quality and Safety in Healthcare (Gothenburg, 2015), can inform similar interventions in other hospitals.
References
Abstract 0156 – E-Poster EPM.084
Improving clinical care of critically unwell patients through development of a simulation programme for use in a Malawian hospital
Freedom from Fistula Foundation
We developed a programme of low-cost simulation-based training for nurses at the Fistula Care Centre, Lilongwe, Malawi with the Freedom from Fistula Foundation. After a successful pilot programme, we repeated the course with pre- and post-course data collection to assess whether the course improved staff confidence in the management of critically unwell patients.
Delegates completed a pre- and post-course questionnaire, using visual analogue scales, to evaluate whether the course had improved their self-reported confidence in various aspects of acute illness management. Areas assessed were: sick patient recognition, ‘ABCDE’ assessment, emergency management, handover, recovery position, airway maneuverers and CPR.
Results were analysed for statistical significance using a paired t-test.
In each area, staff showed a significant increase in self-reported confidence following the course. Across all areas, the mean pre-course confidence was 3.29, increasing to 4.71 post-course, a 43.3% relative increase.
Our course used only basic equipment, but was still able to achieve a significant improvement in staff confidence in the management of critically unwell patients. This curriculum, combined with the training of a local clinician to continue running the course, will make it a sustainable programme that can continue annually. This will facilitate teaching the healthcare staff, who have previously had little or no access to acute care training, to become competent in managing critically unwell patients, which we believe will positively impact patient care in the long run for all women attending the centre.
Changes in self-reported acute illness management abilities pre- and post-course.
Reference
Abstract 0247 – E-Poster EPM.085
Deranged vital signs as prognostic indicators for mortality on an intensive care unit in a low-income country
1University of Birmingham
2Karolinska University Hospital
3Kilimanjaro Christian Medical Centre
4University of Newcastle5Western Sussex Hospitals
Tidal volume (ml/kg of PBW). Gender-specific tidal volume.
Reference
Abstract 0140 – E-Poster EPM.086
A novel approach to patient safety in a Scottish Emergency Department with some help from the African philosophy of Ubuntu
Aberdeen Royal Infirmary
Patient safety and reduction of errors are key in today’s world of medicine. Errors are defined as a failure of a planned action to be completed as intended, or use of a wrong plan to achieve an aim. Not all errors lead to adverse events or harm.
The emergency department (ED) has several characteristics that make it prone to errors, such as high diagnostic uncertainty, high decision density, high cognitive loads, rotational junior medical staff, high levels of interruptions and distractions.
It is said that to get to expert level, where one is able to anticipate potential adverse events and prepare for them in advance, requires at least 10 years’ experience. We recognised that every four months we had a new set of junior doctors coming to our ED, and that the learning curve they go through is very steep.
We wanted a way to capture the errors that didn’t lead to adverse events and share that experience through sharing knowledge.
The African philosophy, Ubuntu, has many definitions, but was summed up by Leymah Gbowee as ‘I am what I am because of who we all are’. We have taken the Ubuntu philosophy, and previous experience in industry, and developed a learning from experience (LFE) process in our ED; the Ubuntu Project. The idea behind the Ubuntu Project is that we as a group can improve by sharing the experiences of individuals. We aim to capture and share learning from errors in a positive, constructive manner.
All members of the ED team are able to contribute to the Ubuntu Project by sharing their learning in person, email or via our department’s website (www.abdn-em.co.uk). Distribution of learning to the entire team is via a document sent out to all ED staff by email as well as hardcopies displayed in staff rooms, workstations, safety briefs and notice boards.
The following topics have been highlighted: blood transfusion, insulin prescribing, non-technical skills (specifically phone communication), oxygen and death and dying.
Reception has been overwhelmingly positive. The hope is that over time it will become normal to talk about errors in order to facilitate learning. The more we can share the safer a place our ED will become both for the patients and the staff.
Abstract 0258 – E-Poster EPM.087
A retrospective cohort study of a Tanzanian intensive care unit
1University of Birmingham
2Karolinska University Hospital
3Kilimanjaro Christian Medical Centre
4University of Newcastle
5Western Sussex Hospitals
Abstract 0083 – E-Poster EPM.088
Non-invasive ventilation in a DGH ICU: The North Middlesex Hospital experience
North Middlesex Hospital
Non-invasive ventilation (NIV) is currently an essential procedure to treat patients with respiratory failure. New ventilators and devices have improved tolerability and safety during its administration. The North Middlesex Hospital (NMH) intensive care unit (ICU) assumed the responsibility of the NIV unit after the Enfield and Haringey strategy in 2013. This study presents a comparison between NIV patients treated currently at NMH ICU (Group A) with a 2014 ICU historical database (Group B).
Variables are presented as mean±SD. Pearson chi-square and Student’s t-test were used for qualitative and quantitative variables, respectively, to compare groups. When the distribution of the sample was not normal (Hospital LOS) a U-Mann-Whitney was employed to compare both groups. The statistical analysis software employed was Systat, Inc. (v 13.1).
Comparison of all NIV patients between both groups.
Comparison of only COPD patients on NIV between both groups.
Abstract 0226 – E-Poster EPM.089
Can we improve how we ventilate our patients in theatre?
University Hospitals Birmingham
Progressive atelectasis is a feature of general anaesthesia. Traditional teaching was that a large tidal volume (10 to 12 ml per kg actual body weight) would recruit collapsed alveoli. It is now understood that cyclical collapse and over distension are deleterious. Lung protective ventilation uses 6 to 8 ml per kg of predicted body weight together with positive end expiratory pressure in an effort to minimise barotrauma, atelectotrauma and shear stresses which contribute to ventilator-induced lung injury.
In patients undergoing major abdominal surgery, the incidence of post-operative pulmonary complications is significantly improved with lung protective ventilation, when compared to non-protective mechanical ventilation. Lung protective ventilation improves clinical outcomes and reduces healthcare utilisation in these patients.
The aim of this audit was to identify the extent to which low tidal volume ventilation is used in patients undergoing general anaesthesia with mechanical ventilation at a university teaching hospital.
The audit was conducted on five normal weekdays between January and March 2016. The intra-operative tidal volume (VT) of 40 patients undergoing general anaesthesia with mechanical ventilation, and the duration of anaesthesia was recorded. Age, gender, ASA grade and operative procedure of each patient were also documented. Predicted body weight (PBW) of each patient was calculated with reference to their height and previously defined gender specific PBW and VT charts. VT expressed as millilitres per kg, was then calculated for each patient.
The recommended tidal volume was delivered in 55% of cases. Forty-five per cent of patients received tidal volumes >8 ml/kg of PBW.
Forty-five per cent of the patients in this audit had suboptimal ventilation during their surgery at a mean duration approaching 3 h. The likelihood of receiving a non-protective ventilation strategy was greater if the patient was female. We hope the implementation of an education package and a PBW tool attached to the anaesthetic machine may ameliorate these effects.
References
Abstract 0064 – E-Poster EPM.090
Driving pressure as a part of lung protective ventilation at a large district general hospital
Royal Cornwall Hospital
Acute respiratory distress syndrome (ARDS) is a syndrome of non-cardiogenic, inflammatory pulmonary oedema which may lead to catastrophic refractory hypoxia. The pathogenesis is not fully explained, but evidence suggests an element of iatrogenesis, with volutrauma, barotrauma, atelectrauma and oxygen therapy all implicated. More recent evidence implicates ‘driving pressure’ (the difference between Ppeak and PEEP) as a cause of ARDS.1 This project aimed to evaluate the lung protective ventilation strategy for our patients with ARDS, particularly whether we comply with recently identified targets for driving pressure.
Single centre retrospective audit of patients diagnosed as ARDS. The unit database was screened for adult patients ventilated for greater than three days between January 2015 and January 2016. Notes and chest imaging were then reviewed to check for bilateral infiltrates consistent with ARDS and to exclude cardiac failure. Patient characteristics, classification into mild/moderate/severe by median PaO2:FiO2 ratio, 6-h ventilatory data, underlying diagnosis, 30-day mortality were all recorded.
A total of 176 patients were ventilated for three days or more over the study period. One hundred and thirty-five were excluded after review of notes and imaging. Therefore, 41 patients were included as having ARDS. The median age was 58; median APACHE-II score was 20. Eighty per cent had primary lung insult, e.g. pneumonia. Six were classified as mild, 32 as moderate, 3 as severe. Overall 30 day mortality was 17%.
Median ventilatory values are shown in Table 1.
Fifty-nine per cent ventilatory measurements deviated beyond tidal volume of 6 ml/kg, 19% beyond peak 30 cmH2O, 60% beyond driving pressure of 13 cmH2O.
Despite this, just seven patients were proned (one mild, four moderate, one severe), with all three severe patients, 13 (41%) moderate patients and nine (17%) mild patients receiving paralysis to aid mechanical ventilation.
Reference
Abstract 0065 – E-Poster EPM.091
How do we measure up? PEEP vs. FiO2 balance in ARDS patients at a large district general hospital
Royal Cornwall Hospital
Acute respiratory distress syndrome (ARDS) is a syndrome of non-cardiogenic, inflammatory pulmonary oedema which may lead to catastrophic refractory hypoxia. The pathogenesis is not fully explained, but evidence suggests an element of iatrogenesis, with volutrauma, barotrauma, atelectrauma and oxygen therapy all implicated. It is probably best practise to balance positive end expiratory pressure (PEEP) against fraction inspired oxygen (FiO2) in order to achieve maximal lung protection whilst managing refractory hypoxia. This project aimed to evaluate our unit’s compliance with these standards.1
Single-centre retrospective audit of patients diagnosed as ARDS. The unit database was screened for adult patients ventilated for greater than three days between January 2015 and January 2016. Notes and chest imaging were then reviewed to check for bilateral infiltrates consistent with ARDS and to exclude cardiac failure. Patient characteristics, classification into mild/moderate/severe by median PaO2:FiO2 ratio, 6-h ventilatory data, underlying diagnosis, 30-day mortality were all recorded. Data were compared to ARDSnet suggested ventilator protocols to assess compliance with PEEP vs. FiO2 tables (Table 1).1
A total of 176 patients were ventilated for three days or more over the study period. One hundred and thirty-five patients were excluded after review of notes and imaging. Therefore, 41 patients were included as having ARDS. The median age was 58. Median APACHE-II score was 20. Eighty per cent had primary lung insult, e.g. pneumonia. Six were classified as mild, 32 as moderate, 3 as severe. Overall 30 day mortality was 17%.
Overall, as demonstrated by data in Table 1, compliance with ARDSnet tables was extremely poor. Overall, those patients on lower FiO2 have too high a PEEP, whereas those on high FiO2 have too low.
FiO2 vs. PEEP matching targets – ‘Low PEEP’.
Percentage patients on appropriate PEEP for FiO2.
Reference
Abstract 0177 – E-Poster EPM.092
Early versus late reintubation in ICU patients
M Konte and
ICU, Xanthi General Hospital
Abstract 0034 – E-Poster EPM.093
Effectiveness of non-pharmacological interventions for delirium management in ICU: Results from a systematic review
Queen's University Belfast
List of non-pharmacological interventions associated with a statistically significant reduction in delirium in ICU.
Abstract 0175 – E-Poster EPM.094
Wide disagreement between alternative assessments of physical function: Patients subjective reports, surrogate reports and smartphone data
1The University of Adelaide
2University of Michigan
3VA Ann Arbor Health System
Premorbid physical function is important when setting goals for critically ill patients. However, surrogate and self-reported estimates are subjective and may be imprecise. It may be possible to extract objective premorbid physical function data from a patient’s smartphone. Specifically, step and global position system (GPS) data might be able to quantify physical function.
We therefore conducted a prospective cohort study to (1) assess the accuracy of surrogate-derived assessment of physical function and (2) determine the feasibility of extracting step and GPS data in survivors of critical illness.
We enrolled consecutively emergency admitted patients over 18 years old, with ICU LoS>48 h, living<50 km from the hospital, who were ambulatory at baseline and owned a smartphone. Sixty-three patients (51 (18) years, 40 men) and their surrogates consented to questioning and 50 patients (52 (18) years, 33 men) consented for smartphone analysis (Figure 1).
Surrogates underestimated patients’ daily steps per day by (median (range)) 1000 (−14000 to 2000) steps but accurately reported time and distance walked prior to requiring a rest (−6 (−460 to 300) min and 0 (−19,800 to 18,000) m) respectively. However, the wide ranges demonstrate the considerable variation at the individual patient level. Similarly, there were relatively weak correlations between surrogate and patient-reported activity (r = 0.35, p = 0.005); r = 0.36, p = 0.004; and r = 0.37, p = 0.003, respectively, for steps, time and distance walked) and Bland-Altham plots showed poor agreement (with limits of agreement of 96% lower to 1170% higher, 97% lower to 1740% higher and 99% lower to 8700% higher for steps, time and distance, respectively).
We extracted step and/or GPS data from 24/50 (48%, 95%CI: 35–62%) phones. In nine phones, the data were from during the hospital admission, as pre-morbid data had been overwritten. Mean step count during the 28-day period prior to admission was 3525 (53 to 12376 steps, n = 13) and mean percentage time spent at home was 56% (23%, n = 11). Associations between objective step count data and either subjective patient or surrogate estimates (r = 0.61 (p = 0.027) and r = 0.63 (p = 0.02) respectively) were stronger than between patients and surrogates’ estimates alone.
These data suggest subjective surrogate estimates of physical activity may be imprecise for individual patients. Obtaining objective pre-morbid data from smartphones was feasible in approximately 50% of patients.
Abstract 0090 – E-Poster EPM.095
Meeting the needs of critical care patients after discharge home: An exploratory qualitative study of patient perspectives
City University of London
The National Institute for Clinical Excellence1 released guidance on a pathway of rehabilitation through critical care to the ward, and the initial recovery at home. Nonetheless, there is evidence of poor uptake of these guidelines, and the support needs of patients and their carers following hospital discharge have been reported as a priority in a recent stakeholder exercise.2 Involving former patients in the design of services, they receive aids compliance and relevance.
Participants were recruited by convenience sampling, through a UK critical care charity and a patient and public involvement group. They took part in a semi-structured telephone interview and were asked about their experiences of support or services they received or required at home. Interview transcripts were analysed using thematic analysis.3
Participants reported a lack of information about what problems they might face once home and sources of help. Participants who were directed to information sources such as booklets and internet sites found them useful and reassuring, and peer support on forums and in support groups was highly valued. It was difficult to access specialist services in the community. Follow-up clinics played an important role in identifying problems, referring to appropriate services and explaining symptoms, but did not always occur at the right timepoint in recovery.
Themes of experiences of support.
References
Abstract 0164 – E-Poster EPM.096
Pre-exisiting level of dependency: The missing element?
Intensive Care Unit, Northwick Park Hospital, Harrow, Middlesex, UK
Reference
Abstract 0190 – E-Poster.097
Impact of prolonged stay on the Cardiothoracic Intensive Care Unit on patient outcomes
1University of Cambridge
2Papworth Hospital
How the association between mortality in ITU changes with the minimum number of days spent in ITU. How the minimum length of time spent in ITU is associated with discharge destination.
References
Abstract 0116 – E-Poster EPM.098
Development and piloting of a screening tool to identify patients with communication difficulties in critical care
Guy's and St Thomas' NHS Foundation Trust
2. To develop and pilot a new communication screening tool on critical care.
References
Abstract 0222 – E-Poster 099
Documentation of patient wishes and goals on critical care
West Suffolk Hospital
Patients with the capacity to make medical decisions may decline treatments offered to them, even where refusal may be considered unwise. Critically ill patients may have capacity during some or all of their admission; however, their acceptance of increasing organ support is commonly assumed rather than explicitly discussed. This may contribute to patients at the end of their life receiving uncomfortable and futile interventions rather than timely palliation. Routine discussion of patient preferences and goals may therefore help to deliver patient-centred care. We audited the recording of patient wishes in their medical records in response to a quality metric document published by our local critical care network.1
We reviewed 20 consecutive patient records for evidence of discussion of their willingness to receive cardiopulmonary resuscitation (CPR), intubation and ventilation (I&V), cardiovascular support (CVS) (vasopressors/inotropes), and renal replacement therapy (RRT). Following this, a mandatory form was introduced where capacitous patients had their wishes to receive these interventions recorded. This form was completed at critical care admission, otherwise as soon as capacity returned. Valid and relevant advance directives were sought and their content recorded for non-capacitous patients. A further 20 consecutive patient records were audited following this intervention.
The results are shown in Table 1. No patient had any recorded preferences for escalation during the baseline audit; only 2/20 had their CPR wishes recorded. This significantly improved post-intervention. Reasons for non-recording were non-capacitous patients or advance directive (4/20); patients where organ support (except CVS) was not offered (1/20); and capacitous patients whose care goals were not sought (5/20).
Frequency of recording patient wishes for escalation in medical records.
Reference
Abstract 0234 – E-Poster EPM.100
Communicating with mechanically ventilated patients: Can using technology help?
1Cambridge University Hospital
2Cambridge University
Difficulty communicating with a patient who is mechanically ventilated is a common clinically encountered problem and is a frequent cause of frustration for both patients and clinicians, in critical care. Multiple studies have highlighted the negative impact of communication impaired by mechanical ventilation.1,2 These clearly highlighted the significant emotional as well as physical distress caused to patients who were mechanically ventilated. Patients themselves report impaired communication as the most distressing experience of critical care as an intervention.3
We surveyed 78 patients in a three-arm enquiry into the quality of patient perception of the quality of attempts to facilitate communication, whilst they were ventilated on a general intensive care unit. In patients identified as CAM-ICU negative and RASS +1/0/−1 (cohort 3), we also conducted a staff survey following 90 interactions to allow comparison of quality perceptions.
Only 33–56% of patients felt they could communicate effectively. Staff over-estimated the quality of interaction; 86% felt communication was effective. High levels of satisfaction were seen with understanding physical needs and levels of discomfort in all groups.
Patients in cohort 3 displayed higher a prevalence of fear (71–83%) and stress (57–83%) than the general ICU population (37% and 40% respectively). The reduced prevalence of fear and stress seen in cohort 3 related to the provision of a specifically designed communication iPad app.
No studies have correlated quality of communication and the prevalence of post-traumatic stress disorder. This assessment has both identified a group that has higher levels of fear and stress and demonstrated a trend in reduction in levels of both.
References
Abstract 0039 – E-Poster EPM.101
NHS Borders Daisy Project: Person-centred care in the ICU
J Aldridge,
Borders General Hospital
The Scottish Government’s Healthcare Quality Strategy1 supports the development of person centredness and improved patient experience. The success of the ‘what matters to me?’ project, led by Jennifer Rodgers at Yorkhill Hospital, prompted us to ask the question ‘do we really know what matters to the patients we care for in the intensive care unit (ICU)’?
The Daisy Project acknowledged the stress that patients and their relatives experience whilst in hospital and aimed to provide truly person-centred care by uncovering and addressing what is important to ICU patients during their hospital stay. It aspired to develop a service that does not assume knowledge on the thoughts of patients, but rather asks them directly and responds to their needs.
The model for improvement2 provided the methodology for the project, using three fundamental questions:
What are we trying to accomplish? How will we know that a change is an improvement? What change can we make that will result in an improvement?
We wanted to give every patient the opportunity to fill in a daisy poster (Figure 1) which would be displayed close to their bed, based on the ‘getting to know me’ form developed by Alzheimer’s Scotland.3 Feedback was collected from questionnaires given to nursing staff before and after implementation of the posters. Patient feedback, when well enough to recall their ICU stay, was also encouraged.
Process date showed success in staff engagement with the project, with numbers of posters being completed each week (Figure 2). Outcome data in the form of staff feedback have been positive. Older patients wrote very specific lists outlining personal interests during their stay, with some describing surprising facts, likes and dislikes which may have gone uncovered had it not been for the initiative.
Daisy Poster. Posters completed per week.
References
Abstract 0172 – E-Poster EPM.102
Barriers to a career in intensive care medicine: A national survey
S Rosling1,
1Nottingham University Hospitals NHS Trust
2Oxford University Hospitals NHS Foundation Trust
3Central Manchester NHS Foundation Trust
4Bolton NHS Foundation Trust
The demand for intensive care medicine (ICM) is growing with a projected 125% increase in demand for critical care between 2013 and 2035.1 ICM is a demanding career that has traditionally been considered an allied specialty in the UK. Current guidelines recommend that ICM is considered a primary speciality with doctors trained accordingly.2
The Intensive Care Society (ICS) trainee committee conducted a survey to assess current opinions and concerns around a career in ICM to identify what factors could be addressed during training and workforce planning to achieve a sustainable career path. The survey was conducted for six weeks prior to closing in August 2015. It was disseminated via email, social media and newsletters to all members of the ICS, RCEM and AAGBI. A total of 656 responses were received. The denominator is difficult to ascertain as it was distributed in several ways. Four hundred and six (62%) of respondents were male; 413 (63%) responses were from ICM consultants and trainees and the remainder from non-ICM trainees (i.e. single-speciality anaesthesia, surgery, medicine, emergency medicine). Consultant workforce gender mix was predominantly male in 507 (98%) responses and predominantly female in eight (2%) responses. The decision to undertake a career in ICM occurred relatively early in more than half of respondents, with over 25% of trainees from anaesthetic background. Reasons for choosing (Figure 1) and concerns (Figure 2) about a career in ICM were ascertained. Five hundred and twenty (79%) respondents felt they had good role models at work. Only 17.4% had been asked about family plans or less-than-full-time (LTFT) training. Twenty-four (4%) respondents worked LTFT with childcare being the commonest reason (>80%). Over 30% did not feel they gained the same training experience when working LTFT. Sixteen (67%) of respondents agreed or strongly agreed that ICM was a good career for LTFT working. This survey has highlighted several key areas that should be addressed to promote recruitment and long-term sustainability in ICM. Work–life balance, sustainability and burnout were the commonest concerns regarding choosing a career in ICM.
Positive findings include the high number of positive role models at work. ICM is a challenging career and the speciality should now focus on addressing how the onerous out-of-hours commitments can be tempered, how women (who make up 55% of medical graduates) can be encouraged to consider it as a career that can work alongside a family, and how psychological support is provided to prevent stress and burnout.
References
Abstract 0118 – E-Poster EPM.103
A semi-structured interview study of intensive care medicine doctors to evaluate their views on learning at night
L Christie
Chelsea and Westminster Hospital
References
Abstract 0066 – E-Poster EPM.104
Designing and implementing a coaching and mentoring programme amongst anaesthetic and critical care trainees at a large DGH
A Harvey, A Pickford, C Pritchett and
Royal Cornwall Hospital
Being a trainee in today’s NHS is a challenging business, with increasing demand, new technology, increased strain and ongoing uncertainty over working conditions. There is increasing recognition that trainee welfare needs confronting and that it is vitally important to develop skills around resilience, stress management, goal setting and task prioritisation alongside clinical work. In business, these skills are often addressed via coaching and mentoring, but within the NHS, this is normally limited to management tiers or as remediation. There is a gap in routine provision for trainees. This project aimed to implement a peer-to-peer coaching and mentoring programme with the long-term aim of improving trainee performance on a variety of measures.
We implemented a programme consisting of monthly sessions for self-selected anaesthetic and critical care trainees, initially based around the Goal-Reality-Options-Will/Wrap model,1 with trainees learning core skills then practising acting as coach or coachee to each other, under the supervision of a trained Executive Coach. A formal supervision session was conducted every four months. Quantitative and qualitative feedback was collected at the end of each session, in order to help shape the programme as it was being implemented.
GROW model.
The programme was well received. Quantitative feedback was positive in all domains (Figure 2).
Results of Post-session surveys.
Themes from the qualitative feedback indicated improvement in problem-solving, development of communication skills and appreciation for a non-threatening place to work through issues and goals.
We have successfully implemented a peer-to-peer coaching programme amongst anaesthetic and critical care trainees at our hospital. This has had wide-ranging benefits for trainees and we are expanding the programme to all trainees at our hospital. We plan to collect data to assess the impact of this on performance, health and wellbeing. We believe that routine coaching and mentoring should be offered as standard to all trainees.
Reference
Abstract 0127 – E-Poster EPM.105
The advanced critical care practitioner role: A workforce solution for ICU? From small acorns…
C Boulanger1, J Platts1,
1National Association of Advanced Critical Care Practitioners
2Advanced Critical Care Practitioner Stepping Hill Hospital
Changes to medical training and increasing scrutiny from the Care Quality Commission on assurance that medical staffing meeting FICM Core Standards make achieving a quality sustainable workforce a key focus.
The National Education and Competence Framework for Advanced Critical Care Practitioners ACCP1 explored how the ACCP role might contribute to anticipated shortfalls in medical staffing. Adopted by the Faculty of Intensive Care Medicine the role has a clear curriculum,2 knowledge and skill and competency requirement. This support cements the ACCP role as a workforce and career option for staff, patients and intensive care.
Since inception in excess of 100 critical care nurses and AHPs have qualified as ACCPs and form part of the medical workforce in ICU. To date, there has not been any evaluation of activity of this role and impact on workforce.
The National Association of Advanced Critical Care Practitioners (NAACCP) undertook a telephone survey of all intensive care units in network regions. A key focus of enquiry was on:
1. Trained ACCPs activity across the UK.
2. Medical rota contribution.
Out of 239 hospital ICU surveyed only 38 (15%) have trained ACCPs currently, the predominance of activity in the North and Scotland. However, of those hospitals ACCPs are contributing up to 262.5 clinical hours per week to the medical rota. Figure 1 shows the relationship between trained ACCP and medical trainee hours for service.
The ratio of medical trainee/ACCP ranged from 26.1 % to 78.1% across the units with ACCPs as part of the workforce. Results show 100 ACCPs practising clinically on the full medical rota across the 24-h period.
This work illustrates what is possible in adoption of the role as a workforce solution. Numbers of units employing ACCPs is currently small; however, those units are seeing an important impact on their medical rota. The number of ACCPs in training has considerably increased with the FICM ACCP curriculum assuring this impact is set to grow across increasing units.
This paper does not explore the quality elements the role brings to the ICU in terms of consistent team members, background critical care experience, impact on patients nor the contribution to the educational experience of trainees on ICU. The previous critical care experience of trained ACCPs is 5–20 years providing an important clinical practice benefit. These aspects warrant further exploration which is the intention of this group.
References
Abstract 0224 – E-Poster EPM.106
“How has the national critical care competency framework impacted patient experience”: The perceptions of service mangers, clinical educators and registered nurses
M Kynaston
Cheshire & Mersey Critical Care Network
Measuring patient experience is a complex area of inquiry, which requires an approach relevant to the patient group under review. For critical care, it is widely acknowledged that additional methods of measuring experience are required, to recognise the unique and often devastating the effects of a life-threatening injury or illness and the wider impact of the critical care environment. A mixed methods approach has been taken to measure current framework adoption rates (national questionnaire), develop a conceptual framework the ‘15 aspects of critical care patient experience’ (systematic literature review) and measure the impact of the competency framework on those identified aspects (regional questionnaire).
Fifteen identified aspects of critical care patient experience:
This analysis identified that 71.7% of critical care service providers across England, Wales and Northern Ireland have adopted the competency framework. Within a representative region, service mangers, clinical educators and registered nurses agreed that the competency framework had supported clinical practice to deliver person-centred care (p < 0.001 in all 15 aspects of critical care patient experience). There was no significant difference seen in the majority of ‘aspects’ between staff group, years of clinical experience or organisation.
Overall responses by aspect of patient experience:
The analyses of the national questionnaire included the number (N) and mean percentage (%) of those providers that have adopted the competency framework. This method was chosen to clearly describe and present the data, as no relationship is being tested. The regional questionnaire has been analysed using SAS v9 software. The data collected from each of the 71 fields (within the 15 identified aspects of patient experience) has been ranked and coded using a Likert-type scale. The chi squared statistical technique was used to investigate the relationship between the dependent (competency framework) and independent (reported effect on patient experience) variables, using the agree vs. disagree responses. The t-test and ANOVA analysis were then applied to explore if there was any statistical significance between responses from different organisations, staff groups or years of clinical experience.
Abstract 0209 – E-Poster EPM.107
An underestimated haematoma
F Khattak1, Z Majid2 and
1Birmingham City Hospital
2Royal Wolverhampton NHS Trust
A 45-year-old male lost control of his motorcycle at 30 mph and collided against a wall with his left side. He complained of pain with marked swelling to the buttock. Reduced sensation in the sciatic nerve distribution was noted on admission and creatinine kinase (CK) was measured at 4883 µ/l (40–320). A CT scan at 26-h post injury demonstrated a large left buttock haematoma, probable sciatic nerve compression and a small un-displaced pelvic fracture. The patient was taken to theatre at 40-h post admission owing to increasing pain and onset of foot weakness. Upon division of the gluteal muscles,, 2.5 L of blood was quickly lost. The wound was packed and the patient transferred for interventional radiology after review by a vascular surgeon. The iliac artery was later successfully embolised.
In our case, the cause of the haematoma was not recognised. Symptoms of sciatic nerve compression were identified early and progressed over a 24-h period; CK was elevated and pain increased despite regular opiates. The elevated CK indicates muscle injury and rhabdomyolysis, which can lead to acute renal failure. Careful attention to hydration and monitoring of renal function is required.
References
Abstract 0203 – E-Poster EPM.108
Tracheomalacia: Chronic obstructive airway disease does not always equal COPD
Royal Alexandria Hospital
Tracheomalacia is a condition characterised by a lack of structural support to the trachea. The trachea is unable to maintain its structure and thus is more likely to collapse during periods of increased extrinsic force. The symptoms can range from chronic cough and wheeze to life-threatening stridor and apnoeic episodes. The vast majority of causes seen are congenital, presenting in the first year of life. Occasionally, however, it can be acquired and present with symptoms of upper airway obstruction, or as was the case in our patient, similar to those of COPD.
We present a case of an intensive care patient, with a previous diagnosis of severe chronic obstructive pulmonary disease, who was found out to instead have tracheobronchomalacia.
The patient went on to have tracheal and main bronchi stenting after which she improved sufficiently to allow extubation.
Of note, had our patient not been admitted to ICU, it is likely that she would have lived the rest of her life with an incomplete label of COPD. The case highlights the need to consider tracheobronchomalacia in cases of COPD with atypical presentations, unusual symptoms or slow weaning from ventilation.
Abstract 0012 – E-Poster EPM.109
A case review of eight patients admitted to the critical care unit with Pneumocystis jirovecii pneumonia
Blackpool Victoria Hospital
Pneumocystis jirovecii (formerly Pneuomcystits carinii) pneumonia (PJP) occurs in immunosuppressed patients and has a high mortality. Causes not secondary to HIV infection are usually related to immunosuppressive therapy.1 The typical course of PJP is a gradual increase in dyspnoea, dry cough and fever.2 Diagnosis is by PCR testing of distal airway specimens. Blood serology is a second line diagnostic tool.2 Treatment for PJP is usually trimethoprim-sulfamethoxazole,3 which usually lasts between 14 and 21 days.4 The use of adjunctive glucocorticoids is recommended for HIV-infected patients with moderate or severe PJP. Glucocorticoid treatment for non-HIV-related PJP is recommended, although based on limited evidence.2,3
We reviewed all patients testing positive for PJP on respiratory specimen PCR who were admitted to an adult intensive care/high dependency unit over a two-year period. Data were collected from the case notes of each patient and entered into a proforma.
We reviewed eight cases; two survived to discharge. We found all eight had an underlying reason for immunosuppression. Methotrexate treatment was associated with four patients. One patient had received radiotherapy and steroids. Two patients were known HIV positive and two patients were newly diagnosed HIV on their admission.
All of our patients were treated with TMP-SMX; one patient was treated with adjunctive steroids.
We found that although there was consideration of an atypical cause for the pneumoni (0–9 days from admittance), investigation for an atypical cause took on average 5.2 days. Diagnosis of PJP occurred in critical care in seven of eight patients. TMP-SMX treatment commenced within 12 h of samples being sent.
Increasing numbers of patients are being exposed to immunosuppressant drugs and recent evidence shows that PJP is also increasing at 7% per year.
Our recommendations are:
1. Immunosuppressive treatment, particularly methotrexate, is poorly recognised as a risk factor for PJP
2. The threshold for testing for HIV remains too high
3. PJP must be considered in any atypical pneumonia
4. In non-HIV cases with PJP, fine, diffuse infiltrates and a diffuse ground glass appearance on chest X-ray are common. An air bronchogram may be visible. High resolution CT may aid diagnosis.
5. Diagnosis was delayed by lack of suspicion prior to critical care admission. Sputum load is typically low, and an awake BAL or a hypertonic saline induced specimen should be attempted if unintubated
6. Consider empirical co-trimoxazole
7. In confirmed PJP patients requiring respiratory support, prescribe glucocorticoids
Abstract 0019 – E-Poster EPM.110
Birds of a feather: An uncommon cause of pneumonia and meningo-encephalitis
S George1,
1George Eliot Hospital
2University Hospital of Coventry & Warwickshire
A fit and well 61-year-old man was admitted to A&E with a one-week history of flu-like symptoms, maculopapular rash, pyrexia and feeling generally unwell. He deteriorated rapidly developing respiratory failure, drowsiness and confusion. He was admitted to the intensive treatment unit (ITU) for respiratory support where his GCS dropped, oxygen requirements increased and he developed acute kidney injury. On day 3, he suffered a convulsive episode.
In the ITU, he required 50–60% of high flow oxygen with intermittent non-invasive ventilation (NIV). He was a keeper of several species of tropical birds, 11 in total, two of which had recently unexpectedly died. A likely diagnosis of community-acquired pneumonia (CAP) with meningo-encephalitis was considered secondary to Psittacosis.
His blood results indicated a CAP, confirmed by computed tomography scanning which indicated a bilateral pneumonia with small pleural effusions. Sputum PCR was positive for Chlamydophila psittacai. CSF showed lymphocytosis and high proteins, but no viral elements were detected. The patient was started on broad spectrum antibiotics (ertepenem and clarithromycin) initially. On day 3, post admission, high-dose macrolide therapy, in the form of doxycycline was started, once the bird history came into light. He improved after this treatment and was later discharged from hospital to complete a three-week course of high-dose doxycycline in the community.
Only 50 cases of psittacosis are reported each year in England and Wales, with less than 1% of these patients requiring ITU. Transmission is through the inhalation of aerosolised bacteria from the faeces, feather dust or respiratory secretions of infected birds. It often presents with flu-like symptoms, being easily dismissed as a simple CAP or influenza. Neurological symptoms include severe headache, photophobia, a reduced GCS and, in rare cases, meningo-encephalitis. Other end organ complications such as endocarditis, renal disease, myocarditis and hepatitis have also been described.
In patients presenting with CAP who have been exposed to birds, psittacosis should be borne in mind. Doxycycline is the choice of drug and should be initiated as soon as infection is suspected. Public health must also be notified.
Abstract 0029 – E-Poster EPM.111
Atypical pneumonia – A case where a newer test improved outcome
D Clarence1,
1Walsall Manor Hospital
2Whiston Hospital
A 57-year-old gentleman presented to A&E with malaise, dyspnoea and rigors. On examination, he was pyrexial with right-sided crepitations on chest auscultation.
Investigations revealed raised inflammatory markers, acute kidney injury, type 1 respiratory failure and right upper lobe consolidation.
He was admitted to the high-dependency unit to commence CPAP as well as piperacillin/tazobactam, clarithromycin and rifampicin, as per trust guidelines for community-acquired pneumonia (CAP). Despite supportive therapy, no significant improvement was observed. On further questioning, he revealed a two-week history of caring for ill pigeons for a relative; on day 4 after admission, throat swabs came back positive for Chlamydophila psittaci by PCR. This prompted an antibiotic change to doxycycline monotherapy and the patient improved significantly within 48 h and was discharged by day 9.
C. psittaci is associated with many bird species and is an under-appreciated atypical cause of pneumonia.1 Symptoms may be mild, non-specific or mistaken for influenza, often leading to a delay in presentation.2
Diagnosis has typically involved serological testing.3 However, these tests are associated with high false-positivity rates and delayed diagnosis.3 PCR has emerged as a superior diagnostic method, offering contemporaneous results, associated with higher sensitivity and specificity rates, playing a significant role in the management of the patient in this case.3
Tetracycline's are considered by some authorities to be superior in the management of psittacosis.1 Less than 1% of correctly treated patients die as a result of C. psittaci infection.
Specific learning points
• Consider psittacosis if flu-like symptoms and history of bird exposure.
• PCR is now available in many hospitals for the diagnosis of C. psittaci and other atypical infections; send throat swabs for PCR early if available in your hospital.
• Based on PCR results, more targeted therapy with alternative agents, such as tetracyclines, can be instituted early.
• De-escalation of atypical cover done on negative results of urine legionella and atypical pneumonia serology alone risks not treating C. psittaci especially in younger patients.
References
Abstract 0139 – E-Poster EPM.112
Multidisciplinary team working in sepsis care: Role of an established “Sepsis Improvement Team” in rapid development of a locally adapted sepsis pathway based on the NICE guidelines (NG51) published in July 2016
Wrightington Wigan and Leigh NHS Foundation Trust
At our organisation, there is a well-established “Sepsis Improvement Team” that meets on a bimonthly basis. The team consist of medical, nursing and managerial representatives from all patient care areas where sepsis is encountered. In addition to facilitating sepsis education Trust wide, this team continuously monitors the performance of sepsis screening, sepsis management, and national CQUIN indicator compliance with ongoing audit.
The sepsis screening tool and sepsis six components in current use are based on the pre-2016 definition of sepsis and sepsis six concepts. But the publication of NICE guidelines in July 2016 meant that these tools needed updating. Our clinical terms were familiar with terms such as SIRS and Sepsis Six, but these were missing in NICE algorithms. Further, unlike previous tools, the NICE guidelines used risk stratification using criteria based on behaviour, respiration, circulation, urine output, skin changes and vulnerability factors.
Existence of a sepsis improvement team facilitated rapid action in preparing a local sepsis pathway. The team met several times and communicated on numerous occasions with this aim. We felt that the NICE algorithms, being of multiple pages and small fonts, appeared too complex for use by multidisciplinary clinical teams on the ground.
Two key suggestions were embraced by all team members.
1. Single page pathways are easier to follow and are more likely to retain compliance
2. Retention of familiar terms would facilitate compliance with the new concepts
Therefore, a single A4 page pathway was prepared. We retained the term “Sepsis Six”, but modified the components to represent the NICE management guidelines. We are delighted to present here the draft versions of these pathways for the benefit of wider intensive care community.
Note: The draft form of the “Adults and 12–17 year old” sepsis pathway is attached along with this abstract. The other age group pathways are being ratified and will be displayed if the ePoster is accepted for the ICS State of the Art Meeting.
Abstract 0145 – E-Poster EPM.113
The future can be old fashioned: Improving sepsis management in the 21st century
J Andrews, Z Broadhead,
Whittington Health NHS Trust
Our aim was to achieve 90% compliance with sepsis pathway completion (recognition), and 90% compliance with timely sepsis 6 and antibiotic administration (treatment).
Our results show significant increases in sepsis pathway completion, Sepsis 6 administration within 1 h of diagnosis, and IV antibiotic administration within 1 h of diagnosis (see attached charts). Notably, timely antibiotic administration has increased from 28% in Q2 2013/14 to 82.2% in Q1 2016/17.
References
Abstract 0169 – E-Poster EPM.114
National Early Warning Score (NEWS) of 3 as a trigger for initiating sepsis screening tool for early identification of sepsis in patients presenting to the emergency department: A prospective observational study
L Gauntlett, K Hall,
Morriston Hospital
Sepsis is common and can potentially kill if treatment is delayed. Sepsis is a time critical condition. Immediate recognition and appropriate treatment is the cornerstone of success. Sepsis, if left untreated, progresses to septic shock. A strong correlation exists between the timing of antibiotic administration and in-hospital mortality.1
NEWS is used throughout NHS as an early warning system to identify acutely unwell patients.2 Routinely, a NEWS score of 4 would trigger sepsis screening.
The aim of this study was to use a NEWS score of 3 to trigger sepsis screening to help clinicians in earlier recognition and treatment of sepsis.
This was a prospective, single centre, observational study in a University Hospital Emergency Department that sees over 90,000 patients annually. All adults (>18 years) with acute presentation between August 2015 to February 2016 were included in the study.
Completed sepsis proforma for all patients were collated. ED records of patients with NEWS ≥3 were reviewed using Zylab (e-Discovery and information management system). We reviewed the hospital notes of patients with a NEWS score of 3 to see if by using this score as a trigger for sepsis, we could identify sicker patients earlier.
Review of the collected sepsis data showed a NEWS of ≥3 in 299 patients. A total of 129/299 patients (43%) had a NEWS of 3 and 170/299 patients (57%) had NEWS >3.
Of the 129 patients, 58 were discharged home, 66 were admitted (HDU-5, AMAU-36, surgical-17 and orthopaedics-6, OMFS-1, gynae-1) and 5 discharged against medical advice (DAMA) or did not wait (DNW).
In this group with NEWS of 3, 35/129 patients (27%) were found to have suspected infections. Tewnty-four patients out of 35 (68.5%) were admitted for treatment and nine patients (25.7%) were discharged home from ED.
By using NEWS score of 3 as a trigger, we were able to identify 35 patients in whom the identification of sepsis and appropriate treatment was initiated at an earlier stage compared to using NEWS of 4 as a cut off.
References
Abstract 0009 – E-Poster EPM.115
Improving the management of sepsis in a district general hospital by implementing the ‘Sepsis 6’ recommendations
Timaru Hospital, South Canterbury District General Hospital
Sepsis is a common condition with a major global impact on healthcare resources and expenditure. The Surviving Sepsis Campaign has been vigorous in promoting internationally recognised pathways to improve the management of septic patients and decrease mortality. However, translating recommendations into practice is a challenging and complex task that requires a multi-faceted approach with sustained engagement from local stakeholders.
Whilst working at a district general hospital in New Zealand, we were concerned by the seemingly inconsistent management of septic patients, often leading to long delays in the initiation of life-saving measures such as antibiotic, fluid and oxygen administration. In our hospital, there were no clear systems, protocols or guidelines in place for identifying and managing septic patients.
We therefore launched the Sepsis 6 resuscitation bundle of care in our hospital in an attempt to raise awareness amongst staff and improve the management of septic patients. We introduced a number of simple low-cost interventions that included educational sessions for junior doctors and nursing staff, as well as posters and modifications to phlebotomy trolleys that acted as visual reminders to implement the Sepsis 6 bundle.
Overall, we found there to a be a steady improvement in the delivery of the Sepsis 6 bundle in septic patients with 63% of patients receiving appropriate care within 1 h, compared to 29% prior to our interventions. However, this did not translate to an improvement in patient mortality.
This project forms part of an ongoing process to instigate a fundamental culture change among local healthcare professionals regarding the management of sepsis. Whilst we have demonstrated improved implementation of the Sepsis 6 bundle, the key challenge remains to ensure that momentum of this project continues and forms a platform for sustainable clinical improvement in the long term.
Abstract 0126 – E-Poster EPM.116
The initial management of patients meeting the SIRS criteria at a large tertiary care teaching hospital
N Gautam and S Vindla
University of Birmingham
Sepsis is the initiation of the systemic inflammatory response syndrome (SIRS) due to infection.1 It is a major cause of patient mortality and morbidity.2 Sepsis may present as physiological changes in various vital signs – the SIRS criteria.3 The UK Sepsis Trust guidelines state that any patient fulfilling two of the SIRS criteria should be screened for sepsis and treated as necessary.4
This service evaluation aimed to retrospectively assess the management of patients meeting the SIRS criteria at a tertiary care hospital. From a larger study involving approximately 30,000 patients, 39 were randomly selected. It was assessed whether: patient notes indicated if infection was considered; any relevant investigations for infection were performed and any antimicrobial treatment was administered. The time at which each of these actions occurred was also assessed to determine the quality of care patients received.
Consideration of infection was documented in the notes in only 48.7% of subjects. Here, infection may have simply not been considered. However, 87% of patients received at least one of the following: infection indicated in the notes, an investigation or antimicrobial treatment. Therefore, it is likely that infection was often considered but poorly documented. The median time to perform an investigation for infection was longer than the time taken to administer antimicrobials (23.5 and 8.4 h, respectively). In addition, fewer patients had investigations performed (56.4%) than antimicrobials prescribed (76.9%). This suggests that antimicrobials were mostly administered empirically. No investigation was performed or antimicrobial prescribed in 13% of patients, most likely because the patient’s clinical presentation suggested that infection was unlikely.
From this study, it is recommended that hospital wards should adopt an alerting system (such as MEWS) to identify patients meeting the SIRS criteria. In addition, a sepsis tool should be made available to help document possible infection and guide management.
References
Abstract 0081 – E-Poster EPM.117
A national survey of UK intensive care medicine (ICM) trainee experience of critically ill obstetric patients
A Quinn1,
1Anaesthetics Department, James Cook University Hospital
2Intensive Care Department, Oxford University Hospitals NHS Foundation Trusts
3Intensive Care Department, Royal Berkshire Hospital NHS Foundation Trust
Ninety-eight per cent of anaesthetic trainees had obstetric anaesthesia experience, but only two EM trainees and no acute medicine trainees had any obstetric experience.
The varying experience of ICM trainees in managing obstetric complications (Figure 1) was reflected in trainees’ self-perceived confidence. Fewer trainees from non-anaesthetic backgrounds agreed they would ‘feel confident’ managing obstetric patients with respiratory failure, sepsis, major haemorrhage or difficult airways. Most non-anaesthetic ICM trainees had never intubated a woman greater than 20 weeks’ gestation.
Only a third of trainees had awareness of local (33.3%) or national (38.6%) guidelines for maternal critical care.
Only 46.8% had received local or regional teaching and 23.2% of trainees had undergone simulation training in maternal critical care.
Most non-anaesthetic ICM trainees do not feel that their training needs in maternal critical care can be met through their current programme (Figure 2).
Eighty-six per cent trainees expressed interest in attending simulation training.
There are designated maternal critical care teaching/simulation days within our region for stage 1/2 ICM teaching, and our forthcoming conference is correspondingly themed.
Emerging multi-disciplinary simulation-based maternal critical care courses facilitate cross-speciality professional interaction, managing complex cases in a safe environment.3 This survey endorses the availability of such courses nationwide.
References
Abstract 0191 – E-Poster EPM.118
The development of an acute multidisciplinary team to enhance intensive care without borders
R Hunt
Derriford Hospital
The acute care team (ACT) is a multidisciplinary team working in a busy teaching hospital with a wealth of clinical skills and experience. The team was developed in 2011 through the combination of the acute and chronic pain teams, the vascular access team, the outreach team and the emergency resuscitation teams. Since 2014, the team have become responsible for the follow-up of patients on the trauma pathway and they assist with intrahospital transfers when needed. The team is available 24 h a day, 365 days a year, and is very active, averaging over 1000 patient interactions a month. They are invaluable in providing timely, safe, quality care to patients in the hospital. Testament to this is that the Care Quality Commission singled out the team in their most recent report, describing the ACT as outstanding.
Emergency care of the unwell, deteriorating patient has become a particular feature of the team's workload since 2012. As seen in Table 1, the number of medical emergency calls has more than doubled, whereas the number of in hospital cardiac arrests has decreased significantly.
The outcomes for patients having an in hospital cardiac arrest have also improved significantly. Survival over the last year has increased to 34% and we are now one of the leading hospitals in the country for survival after in hospital cardiac arrest.
Number of medical emergencies and cardiac arrests by year.
The ACT is under appreciated, yet offers huge benefit to patients and provides invaluable assistance to the busy and very appreciative junior doctor workforce. The development of the doctor's assistant (DA) role has expanded the team over the last two years and there are now 19 DA's within the ACT. This expansion will mean we can provide the services of this outstanding team more effectively and so improve the quality of patient care within our hospital.
Abstract 0253 – E-Poster EPM.119
Nasal high flow oxygen for community-acquired pneumonia
A Ellison, S Ingley and
Glan Clwyd Hospital
Baseline characteristics.
Outcomes.
References
Abstract 0135 – E-Poster EPM.120
Obstetric admissions to intensive care in a medium-sized district general hospital: An eight-year audit
C Brennan1,
1Russells Hall Hospital
2University of Birmingham
The proportion of patients admitted with haemorrhage (40%), pre-eclampsia/eclampsia (16%) and anaphylaxis (6%) was higher than the national average, whereas non-obstetric reasons (20%) were lower. No admissions for venous-thromboembolism were recorded and sepsis-related admissions were similar to the national average (4% compared with 3.1%–4.9%). Reasons for these discrepancies may be assessed to support future improvements to care.
Seventy per cent received advanced respiratory and 26% received advanced cardiovascular support, of whom the median length of organ support was 12 and 13 h, respectively. Median APACHE II scores in these patients were higher (17 and 17.5, respectively) compared to the overall median (16 and 16 h, respectively). Certain admissions also had higher organ support requirements, for example amniotic fluid embolism and post-haemorrhage – see Figure 1.
Demographic and outcome data (local/national).
References
Abstract 0120 – E-Poster EPM.121
Magnetic stimulation ventilates apnoeic patients under general anaesthetic
M Grey1,
1School of Sport, Exercise and Rehabilitation Sciences, University of Birmingham
2Department of Anaesthesia and Critical Care, University Hospitals Birmingham NHS Foundation Trust
3Department of Critical Care Medicine, Queen Elizabeth Hospital
4Institute for Inflammation and Ageing, University of Birmingham
Current ventilation strategies place the diaphragm in a state of mechanical silence, vulnerable to disuse atrophy.1 Animal studies suggest that even small periods of diaphragmatic work can protect diaphragm fibres from mechanical ventilation induced atrophy,2 but spontaneous breathing is not always feasible during mandatory ventilation. Magnetic stimulation of the phrenic nerves produces contraction of the diaphragm and has been applied in awake volunteers to produce adequate tidal ventilation.3 This has never been performed in patients.
There were two phases: (1) calibration phase with awake healthy volunteers and (2) patients undergoing GA peripheral day case surgery.
We measured resting tidal volume transdiaphragmatic pressure (Pdi) in 11 healthy volunteers and constructed frequency/force curves, using a Magstim Rapid2 (Magstim, Dyfed, Wales) stimulator with two 60 mm figure of eight coils placed over each phrenic nerve. Having produced tidal ventilation in healthy subjects, we used data from the frequency/force curves to perform 1 min of magnetically stimulated ventilation in 10 apnoeic patients under general anaesthetic.
Power/force curves by frequency. Patient demographics and measured parameters during magnetic ventilation.
To our knowledge, this is the first study to demonstrate that it is possible to magnetically ventilate apnoeic patients with an adequate tidal volume. This appears to be safe and effective has the potential for therapeutic benefit in the diaphragm atrophy that occurs in ventilated ITU patients. We plan to investigate its use further in this setting.
References
Abstract 0086 – E-Poster EPM.122
A novel method for identification of patients at risk of deterioration using FACS
1North Middlesex Hospital, Intensive Care Unit
2Sheffield Hallam University, GMPR Group
Experimental group (EG): Patients at risk of deterioration who agreed to participate in the pilot study on facial analysis in critical illness (HRA approved, IRAS 165739). After taking consent from the patients a 5-min video was recorded, encrypted and analysed later using FACS by a trained psychologist. The micro-expressions (or AU) were assessed for the upper face (UF), head position (HP), eyes position (EP), lips and jaw position (LJ) and lower face (LF). Variables are presented as mean±SD. The Pearson chi-square test was employed for inferential analysis (Systat, Inc. (v13.1) was used.
The EG (N = 4, five videos) was formed of 2 females and 2 males with an age of 63.4±3.2, the NEWS score was 6±4.3 and the pain score was zero during the recording of the video. The most frequently detected AU over multiple video frames were AU43 (80%) and AU15 (100%).
The comparison of AU43 and AU15 between both groups was not statistically significant.
AU lower face. AU upper face.
References
Abstract 0121 – E-Poster EPM.123
Pharmacokinetics of paracetamol in critically ill patients: A prospective observational study
J Baldwin, A Herbert,
Lancashire Teaching Hospitals NHS Foundation Trust
In healthy patients at the end of a 15-min infusion of 1 g IV paracetamol, the maximal plasma concentration is approximately 30 mg/l. Analgesic effect commences within 5 min of infusion, peaks within 1 h and lasts 4–6 h. Antipyrexial activity typically reduces fever within 30 min, and lasts around 6 h. Plasma concentrations of paracetamol between 10 and 20 mg/l have been shown to produce antipyretic effect but the concentrations to produce analgesia are not well defined.3
We aimed to determine whether paracetamol pharmacokinetics in critically ill patients differs from the currently accepted mechanisms in healthy individuals.
Serum paracetamol levels were measured prior to administration of IV paracetamol, and then at 15 (end of infusion), 30, 60, 120, and 180 min post-infusion. Levels were taken again 18 h post-dose (5 half-lives), and once daily on each of the following two days. Temperature, blood pressure, C-reactive protein level and neutrophil count were also recorded, along with reason for critical care admission, APACHE II and ICNARC scores.
References
Abstract 0244 – E-Poster EPM.124
Fibrinogen replacement in major haemorrhage: A computer generated model
G McCarthy1,
1Belfast City Hospital
2Craigavon Area Hospital
3Instituto Nacional de Pediatría
shows the fibrinogen concentration as mean and 95% CI in each group along with the percentage of subjects with a fibrinogen less than 1.5 g/l. Table 2
References
Abstract 0117 – E-Poster EPM.125
Aspirin in patients with septic shock: A multicentre observational study
C Boerma1, A Boyle2, S Calcinaro3, E Damiani4, R Domizi3, A Donati4, D Donnelly2, A Kuenzel2,
1Medisch Centrum Leeuwarden
2Regional Intensive Care Unit, Royal Victoria Hospital
3University College London NHS Foundation Trust
4Ospedali Riuniti Umberto I Salesi – Lancisi di Ancona
5St Thomas’ Hospital
Results of multi-variate analysis of ICU mortality.
Reference
Abstract 0059 – E-Poster EPM.126
Regional citrate anti-coagulation: A retrospective observational study
C Booth1, T Jordan1 and
1Salford Royal NHS Foundation Trust
2University of Manchester
The most common form of continuous renal replacement therapy in the ICU setting is continuous veno-venous haemofiltration (CVVH). For many years, the most common form of anti-coagulation for CVVH was systemic unfractionated heparin (UFH), but recently regional citrate anti-coagulation (RCA) has become more popular. RCA causes chelation of ionised calcium (iCa) inside the extra-corporeal circuit and iCa is a vital component of the clotting cascade. Thus, the chelation of iCa helps to prevent clotting of the haemofilter.
We conducted a retrospective observational study comparing two CVVH protocols used in the Critical Care Unit of Salford Royal NHS Foundation Trust. The first protocol (pre-RCA) used UFH as the anti-coagulant, whereas the second protocol (post-RCA) used RCA. Our study included of 19 patients in the pre-RCA cohort, who received 27 episodes of CVVH, using 53 filters during a four-month period, and 16 patients in our post-RCA group, who received 24 episodes of CVVH, using 38 filters during a three-month period.
The baseline characteristics of both cohorts were similar. We found that median filter lifespan was significantly greater (p < 0.001) in the post-RCA group (33.6 h) when compared to the pre-RCA group (26.6 h). There was no significant difference in the lifespan of the clotted filters, with a median lifespan of 30.4 h in the pre-RCA group, and 33.6 h in the post-RCA group (p = 0.731). The rate of filter survival was also significantly greater in the post-RCA group (p < 0.001).
There was significant disparity in the reasons for filter termination between the two protocols. In the pre-RCA group, the most common reason for termination was clotting of the filter (56.8%). Whereas in the post-RCA group, the rates of filter clotting were significantly less (20%) (p = 0.001). Efficacy was similar between the two protocols, with few significant differences between electrolyte levels during the first 36 h on the filter. However, the post-RCA protocol was associated with a lower potassium concentration over the first 36 h, although it still remained within the normal range.
Our results indicate that the new post-RCA protocol is associated with an increase in filter survival time, and this is likely due to decreased rates of filter clotting. Our secondary data indicate that the efficacy of both protocols is similar, although the data were often incomplete and therefore needs to be interpreted with caution.
Kaplan-Meier analysis of filter lifespan.
Abstract 0113 – E-Poster EPM.127
Evaluation of novel protocol safety post implementation of regional citrate anticoagulation haemofiltration on a general/neurosurgical critical care unit
1Salford Royal NHS Foundation Trust
2University of Manchester
All current renal replacement therapy modalities within critical care involve the use of a pump-driven extracorporeal veno-venous circuit. To improve efficiency of these systems, continuous anticoagulation, usually unfractionated heparin (UFH) is required in order to reduce the risk of circuit clotting.
Regional citrate anticoagulation (RCA) offers an alternative to UFH, functioning by chelating ionised calcium whilst the blood is in the circuit, inhibiting the activation and propagation of the coagulation cascade. Infusion of calcium into the blood prior to return to the patient prevents systemic hypocalcaemia. Available literature demonstrates success in safety and efficiency.1
The main complications of RCA are systemic hypocalcaemia, metabolic alkalosis, citrate accumulation and haemorrhage. In addition, the use of citrate containing fluids carries additional cost as these fluids are currently more expensive than standard replacement fluid. Currently available protocols utilise large volume citrate administration which may lead to high levels of associated complications.
We developed a new protocol evolved from previous work by Collin2, reducing the total amount of citrate administered with the aim of reducing associated complications and cost. The protocol utilises continuous veno-veno haemofiltration (CVVH) as the treatment modality dosed at 25 ml/kg/h of ideal bodyweight.
Following the introduction of the new protocol, we evaluated its safety. Patients’ electronic health records, point of care blood gases and laboratory venous blood samples were retrospectively examined for episodes of hypocalcaemia (serum iCa <0.9 mmol/l), metabolic alkalosis (serum HCO3 >35 or pH >7.55), citrate accumulation (iCa:total Ca ratio >2.5) and significant episodes of bleeding.
We evaluated the filtration episodes of 16 patients, using 34 filters over a three-month period totalling 1440 h of haemofiltration. One occurrence of metabolic alkalosis was identified (absolute highest value: 7.58), four occurrences of hypocalcaemia (absolute lowest value: 0.82 mmol/l) none of which were found to have symptomatic or physiological consequence. There were no episodes of citrate accumulation or bleeding complications requiring discontinuation of therapy.
The low rate of metabolic alkalosis and absence of episodes of citrate accumulation may be linked the reduced citrate load delivered by our protocol in comparison to established protocols. In addition, the use of ideal bodyweight dosing also reduces citrate load leading to reduced complications.
Our data suggest that our RCA protocol utilising a CVVH modality has a low complication rate, and may offer clinical and cost advantage through minimising citrate load, provided there is strict protocol adherence and vigilance around potential accumulation and metabolic or electrolyte derangement.
References
Abstract 0134 – E-Poster EPM.128
Economic comparison of citrate-based regional anticoagulation protocols for continuous renal replacement therapy in the intensive care unit
Salford Royal NHS Foundation Trust
The use of citrate for regional anticoagulation (CRA) for renal replacement therapy (RRT) is recommended in the 2012 KDIGO guidelines for Acute Kidney Injury.1 Our intensive care unit delivered a planned change from unfractionated heparin to CRA in 2016.
During design of a new protocol for CRA, cost considerations were a major factor. The costs of RRT are filter consumables, fluids and anticoagulation. Studies have demonstrated increased filter life in comparison to heparin-based systems resulting in a potential cost saving. This is offset by the increased cost of citrate-based fluids required for CRA.
Our ICU uses the Gambro Prismaflex system in CVVH mode. CRA in this system requires the administration of PrismaCitrate 18/0 as predilution followed by administration of calcium gluconate post filter. The dose of RRT provided is delivered in part by the infusion of Prismacitrate 18/0 and in part by the infusion of a replacement fluid post filter (Prismasol 4).
The available protocols for CRA for the prismaflex system are the Collin (CVVH)2 and Kalmar (CVVHDF).3 Both protocols use variable high dose rates (30–40 ml/kg/h) dependent on patient weight. Established practice on our ICU is to provide CVVH at 25 ml/kg/h of ideal bodyweight. We designed a protocol based on Collin’s work in order to provide standardised dose across all weight ranges. During development, we sought to minimise citrate fluid administration with the aim of reducing cost and associated clinical complications (citrate accumulation, hypocalcaemia, metabolic alkalosis).
An economic modelling exercise was undertaken to examine the potential costs associated with each protocol. We compared four protocols:
• existing heparin-based CVVH
• proposed protocol (CVVH)
• Collin (CVVH)
• Kalmar (CVVHDF)
Costs included were:
• Heparin
• Citrate fluid
• Replacement fluid
• Dialysate fluid
• Calcium
• Filter cost
The lifespan of filter circuits was estimated as 20 h for heparin based systems and 40 h for citrate based systems based on local audit and available literature.
Cumulative costs per hour were calculated for each of the described protocols at six patient bodyweights. The median cumulative hourly cost was calculated for all bodyweights across a 72-h RRT period giving an hourly cost of RRT. Hourly costs were: heparin £10.20, proposed citrate £9.22, Collin £9.92, Kalmar £11.90.
This modelling shows that the proposed citrate-based CVVH protocol is theoretically cost saving in comparison to established protocols for citrate and heparin-based RRT. Cost saving is delivered through increased filter life and reduced citrate flow rates.
References
Abstract 0115 – E-Poster EPM.129
Terlipressin for prevention of intra-dialytic hypotension
1Singhealth Anaesthesiology Residency Program
2Department of Anaesthesiology, Singapore General Hospital
Intra-dialytic hypotension (IDH) is a major complication seen during haemodialysis (HD) and is associated with significant morbidity and mortality. The aetiology of IDH is thought to be multifactorial. One postulated mechanism is that there is a lack of rise in arginine vasopressin (AVP) levels during HD despite decreases in blood pressure and blood volume, which are physiological stimuli for AVP secretion.1 Intravenous (IV) vasopressin and intranasal (IN) desmopressin have been used to prevent IDH.2 We describe the administration of terlipressin to prevent IDH in this case report.
Our patient was a 65-year-old Chinese gentleman with multiple medical problems including ischaemic heart disease status post coronary bypass surgery and end-stage renal failure. He was not able to tolerate intermittent HD (IHD) due to IDH despite medical optimisation and had to be switched to peritoneal dialysis (PD). Extensive investigation excluded adrenal insufficiency, autonomic dysfunction and myocardial systolic failure.
He underwent elective adjustment of his PD catheter but suffered haemorrhagic complications and subsequent intestinal obstruction requiring open abdominal surgery. He also developed type 2 non-ST elevation myocardial infarction and was transferred to the intensive care unit. He required noradrenaline and vasopressin infusions for haemodynamic support and was treated with broad-spectrum antimicrobial therapy for septic shock. He received continuous vena-venous haemodiafiltration for three days before he was successfully extubated. He was transitioned to slow low efficiency dialysis (SLED) but required either noradrenaline or dopamine infusions due to IDH. Midodrine and fludrocortisone were ineffective. As such, we prescribed IV Terlipressin 1 mg prior to dialysis and achieved net negative fluid balance without need for additional vasopressors. He was later discharged to the high-dependency unit where he received two successful sessions of ArrtPlus HDF without IDH using the same strategy. Unfortunately, his clinical course was further complicated by PD peritonitis and respiratory failure. He requested for palliation, refused further intervention and eventually passed away.
In this case, we have shown effective use of IV terlipressin in preventing IDH. Terlipressin is a non-selective AVP analogue. It may be preferable in prevention of IDH compared to vasopressin because Terlipressin is easily administered as a single dose via a peripheral IV cannula prior to initiation of IHD, unlike vasopressin which has to be administered continuously via a central venous catheter. Terlipressin may also be preferable to desmopressin because it has mainly vasoconstrictive effect3 unlike desmopressin which exerts mainly antidiuretic action.
References
Abstract 0008 – E-Poster EPM.130
Improving timely medical reviews for patients discharged from intensive care
P Kumar
Timaru Hospital, South Canterbury DHB
Transferring patients from the intensive care unit (ICU) to a general ward is commonly associated with error and adverse events, and is one of the most challenging and high-risk transitions of care. Patients discharged from ICUs often require sustained intensive multi-disciplinary team input, part of which can be provided by nurse or clinician-led outreach teams. Unfortunately, due to a lack of resources, many institutions do not have such programmes. We work in one such hospital with no ICU outreach service for recently discharged patients.
We noted that a disproportionate number of patients recently discharged from the ICU needed acute medical reviews by on-call evening and overnight junior doctors. Furthermore, we noted that many of these patients had not been reviewed by their medical team after having arrived onto the general ward from the ICU.
We aimed to foster a fundamental culture change within junior doctors to review patients within 6 h of arrival onto a ward from the ICU. We introduced simple and low-cost interventions that included educational sessions for junior doctors and ward-based nurses, as well as posters that acted as visual reminders in relevant departments.
Overall, the number of patients discharged from the ICU to general wards that were reviewed within 6 h improved from 22% to 70% in the space of six months. In the same period, the number of patients requiring an acute medical review by the evening or overnight on-call junior doctor dropped from 14% to 0%.
Whilst our project is not necessarily appropriate for many larger institutions that already have outreach teams in place, it is certainly applicable to other similar sized smaller hospitals. We hope that others who face the same inherent barriers are inspired to implement similar projects, to bring about positive change, and ultimately improve the safety of their patients.
Abstract 0023 – E-Poster EPM.131
CT scan requests on intensive care patients: Are they always justified?
P Gruber and
Royal Marsden Hospital
CT scans requested on intensive care (ITU) patients and often involve transfer of unstable patients from the ITU to the radiology department, causing potential transfer risks and interruptions in clinical care. Furthermore, CT imaging exposes patients to ionising radiation and is costly. There is little research into the impact of CT scans on cost, patient safety and clinical decision making.
CT scan requests from a specialist oncology ITU (Royal Marsden Hospital, United Kingdom) over a 12-month period were reviewed (1 April 2014–31 March 2015). Indications for CT scans; requesting clinician; time of scan; organ support required; whether the scan was positive/negative; consequent change in management were obtained using our electronic documentation systems (ICIP, EPR). Radiation absorbed by patients per scan was acquired by reviewing the CT protocol documented on PACs. Estimates of costs were obtained from the radiology department costing records.
A total of 147 CT scans were requested on 96 individual patients. Seventy per cent (n = 103) were performed within working hours (Mon-Fri 09:00–17:00). There were no major adverse transfer events reported and only 18% (n = 26) of ITU patients required ventilator and inotropic support. Most common CT request (29%; n = 43) was to investigate presence of a collection (thoracic/ abdominal/pelvic). Most common scan performed was CT abdomen/pelvis (33%; n = 47). Sixty four per cent (n = 94) of CT scans were repeated within four weeks of a previously documented scan, and 30% (n = 44) of all scans were performed within seven days of a previous scan. Sixty per cent (n = 88) of CT scans showed a clinically relevant positive finding and 51% (n = 75) resulted in direct change in clinical management. Total cost of CT scans over this period was £27,287 (average £285 per patient). Total radiation absorbed from CT scans averaged 935 mGy-cm per patient (range: 50–6845 mGy cm).
CT scanning is safe for patients at this institution – majority of CT scans were performed during working hours; level of organ support required for patients being transferred was low for most patients; none resulted in a major adverse transfer event. Common indications and requests for CT scans reflect the nature of patients admitted to the unit (commonly post-operative and/or immunosuppressed patients). A significant number of CT scans were repeated within a month of a previous scan. There is no comparable cost or radiation data available internally or from other units. We have demonstrated here that CT scans from this department are clinically justified, with over half having a direct impact on patient management.
Abstract 0076 – E-Poster EPM.132
Send the samples – Get the answers
Critical Care Department, Aintree University Hospital NHS Foundation Trust
References
Abstract 0108 – E-Poster EPM.133
Point-of-care access to guidelines in the critical care environment
1Freeman Hospital, Newcastle upon Tyne Hospitals NHS Foundation Trust
2Royal Victoria Infirmary, Newcastle upon Tyne Hospitals NHS Foundation Trust
The delivery of consistent, high-quality care for patients may be improved by clinical practice guidelines (CPGs).1 Critical care is an increasingly complex and information heavy environment where CPGs aim to reduce unwarranted variations in practice. Access to CPGs at the point of care has been highlighted as key to CPG implementation.2
To assess the current guideline delivery system in our hospital, clinical staff were asked to log in, locate, access and extract set information from two CPGs using the current web-based intranet. Each stage of the process was timed. Users were asked to rate both the system and guidelines for ease of use, and give relevant feedback.
We timed 11 clinicians, 7 doctors and 4 senior nurses or advanced nursing practitioners. The median time to log in and access the guideline search page was 55 s. The median time from logging in to location of CPG1 and CPG2 was 300 and 345 s, and to answer the clinical questions was 345 and 394 s respectively. Locating CPGs was rated as difficult or very difficult by 73% (8) for CPG1 and 100% (11) for CPG2. Once a CPG was found, answering the clinical question was only rated as difficult by three users in total, with no one reporting very difficult. Free text feedback stated that the computer logon process was unnecessarily slow and that the web-based search function was not sufficiently powerful. Clinicians asked for an improved search function that was easy to use, had better keywords, searched within documents, and would take you to the specific position in the guideline rather than to the frontpage.
It is clear that end users expect intranet guidelines system to have a powerful, easy to use search function that performs as well as the ubiquitous internet search engine. Simple, fast access to up-to-date CPGs gives timely decision support for clinicians, and aids the delivery of standardised, high quality, best practice care for critically ill patients. Current barriers including slow infrastructure, poor search engines and large documents reduce the engagement of clinicians and jeopardise effective implementation of CPGs.
In response to these issues, we are exploring methods to improve guideline indexing and search mechanisms within our intranet. We have also won funding for a pilot of the Ignaz cross-platform mobile app for guidelines3 with a subset of critical care guidelines. This allows true point-of-care access which we hope will significantly improve CPG accessibility and engagement.
Impact of CAP on ITU admissions, bed days and mortality. ITU admissions and distribution of the two most commonly identified pathogens.
References
Abstract 0257 – E-Poster EPM.134
Benchmarking against the Guidelines for the Provision of Intensive Care Services (GPICS): Development of an Excel tool for ICUs
T Anderson1,
1Hospitals NHS Foundation Trust
2Intensive Care Unit, Freeman Hospital
In April 2015, the Faculty of Intensive Care Medicine (FICM) and the Intensive Care Society (ICS) published the first edition of guidelines for the Provision of Intensive Care Services (GPICS). These guidelines were the first step towards the development of a definitive reference source for the planning and delivery of UK Intensive Care Services. However, as clinicians, we found it difficult to benchmark our ICU’s structure and clinical practices to the several hundred standards and recommendations included within the 60 sections and 5 chapters of GPICS. In particular, it was difficult to get an overview of the guidelines as a whole and identify areas in need of improvement. We set out to develop an Excel tool to facilitate the benchmarking process.
We copied all the standards and recommendations from the GPICS document verbatim into an Excel file, preserving the chapter headings, subheadings and section titles. Next to each standard or recommendation we placed a Red-Amber-Green drop down box, with an option of “Fully met”, “Partly met” or “Not met” (Figure 1). We inserted hyperlinks to allow easy navigation between sections and a PDF writer to generate a one-page summary report (Figure 2). Working with senior clinical staff, allied health professionals and clinical managers, we then self-assessed our ICU against GPICS using the tool.
While the process confirmed areas where we were already aware of a gap between our service and national guidance (rehabilitation after critical illness), it also identified new areas in need of development. The one-page summary report provided an at-a-glance overview of our ICU’s structure and clinical practices when benchmarked against national guidance.
In summary, the Excel tool provided a simple tool for an individual ICU to benchmark itself against the large number of standards and recommendations in GPICS, and identify priority areas for improvement.
Data entry using the Red-Amber-Green system for each standard or recommendation. One-page summary
Abstract 0261 – E-Poster EPM.135
Referrals to intensive care: An analysis of a UK trainee database
1Imperial College
2Hammersmith Hospital, Imperial College Healthcare NHS Trust
Several factors determine the outcome of critically ill patients. The process of healthcare delivery plays an important role, and deserves particular attention, as it is often modifiable. For all but specialist intensive care units (ICUs), emergency referrals represent a significant workload, regardless of whether or not the patient referred is ultimately admitted to an ICU. There is limited/no data available regarding both the nature and amount of work ICU junior doctors perform outside of their ICUs in responding to emergency referrals.
An online logbook, “MedELogbook” (www.medelogbook.com) designed for ICU junior doctors. Although designed primarily as an aide for training, it provides a novel approach to obtaining an insight into the working patterns of ICU junior doctors.
The objective of this study was to explore the pattern of referrals by time of day/day of week and to ascertain any association between time of day/day of week and factors that may be associated with workload (time taken to see referral) or decision making (discussion with consultant, and decision to admit) of ICU junior doctors in the UK.
A total of 2476 referrals were entered by over 100 trainees in the UK. The greatest frequency of referrals to ICU is during weekdays in hours. There is no association between time of day/day of week and time taken to see referrals, ICU admissions, or resuscitation status of patients. Almost half of all referrals were made in-hours during weekdays, with the highest referral rates between 08:00–19:00 and 21:00–23: 00 h. This reflects the ward staffing levels and A+E attendances, respectively. However, we found no association between time of day/day of week and decision to admit patients to the ICU, despite the fact there were fewer consultant discussions out of hours. Eight per cent of patients admitted to the ICU had no active decision about resuscitation status, and trainees spent less time with these patients. Such information is crucial for workforce planning among ICU junior doctors, as the distribution of trainee workforce needs to reflect trainees’ workload.
Variation in the number of referrals seen by time of day. There is a bimodal ‘surge’ in the number of referrals to ICU at 08:00–10:00 and again at 21:00–22:00 h on weekdays. The number of referrals made over weekends demonstrated less variability over 24 h. Proportions of referrals made by time of day/day of week and location of referral.
Abstract 0141 – E-Poster EPM.136
Access to MRI for patients requiring mechanical ventilation – A UK-wide survey
T Pain,
Royal Surrey County Hospital NHS Foundation Trust
Reference
Abstract 0211 – E-Poster EPM.137
Emergency intubation in the intensive care unit and the emergency department in a UK teaching hospital
J Holden, E Hoogenboom,
Royal London Hospital
All patients were preoxygenated and 54% (88) ventilated prior to intubation: 34% (55) with a mapleson circuit, 10% (16) with BVM, 3% (5) with NIV, 1% (2) with an LMA. Additional supplementary oxygenation with nasal cannulae was used in 6% (10) patients.
Planning for intubation using a checklist occurred in 35% (58), difficult laryngoscopy was predicted in 18% (30), and formally assessed in 44% (73) patients. Seventy per cent (115) intubations were grade I, 25% (41) grade II, 4% (6) grade III and 1% (2) grade IV. Overall successful intubation at first attempt was achieved in 88% (144) of cases, second attempt 11% (18) and third attempt 1% (2) (Figure 1). Of those with predicted difficult laryngoscopy, only 27% (8) required more than one attempt. Most intubations were confirmed using waveform capnography 92% (151).
Experienced doctors (>100 intubations) carried out 84% (137) of intubations, all second and third attempts were carried out by experienced doctors. The standard Macintosh laryngoscope was used for all but three intubations; the Airtraq or C-MAC was used on subsequent attempts. A bougie was used in 37% (61) intubations.
Propofol (0.2–4 mg/kg, median: 1.1 mg/kg). Fentanyl (0.8–6.6 µg/kg, median: 2.0 µg/kg) and rocuronium (0.1–2 mg/kg, median: 1.1 mg/kg) were the most frequently used induction agents in 74% (121), 68% (111) and 78% (128) of intubations, respectively.
Hypotension requiring vasopressors was the most common complication (Figure 2).
Further work will aim at improving the planning of emergency intubation and pre-intubation checklist utilisation.
References
Abstract 0236 – E-Poster EPM.138
Correlation between incidence of VAP and reintubation rate in ICU patients
K Pagioulas, M Plassati and A Vakalos
Xanthi General Hospital
Abstract 0043 – E-Poster EPM.139
The content of plasma selenium in early admitted septic patients: A brief report in Iranian population
1Iran University of Medical Sciences
2Isfahan University of Medical Sciences
Abstract 0152 – E-Poster EPM.140
Procalcitonin clearance as a prognostic indicator in critical care patients with severe sepsis
Russells Hall Hospital
Procalcitonin (PCT) is the hormonal precursor of calcitonin. Studies have shown elevated levels of serum calcitonin to correlate positively with presence of bacterial sepsis resulting in its use as a biomarker of sepsis and to distinguish it from other inflammatory pathologies causing an SIRS.
At Russells Hall Hospital Critical Care Unit since 2011, it became established practice to measure daily PCT measurements on patients treated with antibiotics for sepsis. This has allowed us to effectively identify those patients with bacterial sepsis and distinguish them from those with SIRS alone. It has also aided in the de-escalation of antimicrobial therapy in patients with resolving sepsis, reducing overall antibiotic days on the critical care unit.
Recently, some smaller studies of patients with severe sepsis have found a link between the tendency of the plasma concentration and clearance of PCT during the first 24–48 h of admission, and overall prognosis.
• To identify whether clearance of serum PCT predicts better outcomes in patients with severe sepsis
• To determine if PCT clearance can be used as a predictor not only in the early period (24–48 h post admission), but over a longer duration (>6 days)
Median PCT of the first three days (M1) and the Median of the latter three days (M2) over six-day period were calculated and clearance estimated.
Patients were grouped into two categories based on the clearance of PCT; those who demonstrated <30% decrease in PCT over a six-day period, and those who demonstrated >30% decrease in PCT over a six-day period. This was compared to mortality data from the two groups
In patients on CVVH who demonstrated a PCT clearance <30%, the relative risk of mortality was twice that of non-CVVH patients (RR: 4.33).
PCT predicts poorer outcome when it is cleared slowly, and this prediction amplified in sicker patients despite being on RRT.
References
Abstract 0074 – E.Poster EPM.141
Clinical outcomes with haemofiltration in septic shock
Scunthorpe Hospital
Sepsis is one of the common indications for admission to ICU. The use of haemofiltration in sepsis has become a common modality for the management of septic patients.
The aim of our study is to compare the outcomes of the patients admitted to our ICU with septic shock who received continuous veno-venous haemofiltration (CVVH) as part of their treatment, in comparison to those with septic shock who did not receive CVVH.
This was a retrospective study that included all patients admitted to our ICU with septic shock in the year 2015.
The age, sex, ICNARC, APACHE2 and mortality prediction scores were compared between the two groups. There was no significant difference between the two groups in regards to age and sex (p = 0.8185, 0.5229, respectively). However, the patients who received CVVH had significantly higher ICNARC, APACHE2 and mortality prediction scores (p = 0.0232, 0.0001, 0.0200, respectively).
Outcomes that were assessed in this study were 28-day and six-month mortality, duration of hospital stay, improvement in renal function, improvement in inflammatory markers, inotropic requirement and development of chronic kidney disease.
There was no significant difference between the two groups in regards to 28-day and six-month mortality, duration of hospital stay or development of CKD (p = 0.4017, 0.4000, 0.5822, 1.0000 respectively). There was also no significant difference in the improvement of inflammatory markers (p = 0.1067 for WCC and p = 0.5393 for CRP).
However, patients who received CVVH had a significantly better improvement in their renal function (p = 0.0001 for urea, p = 0.0001 for creatinine). It was also found that patients who received CVVH as part of their treatment for septic shock required significantly more inotropic support than those who did not receive CVVH (p = 0.0075).
In conclusion, our study has demonstrated that although the patients who received CVVH had significantly higher ICNARC, APACHE2 and mortality prediction scores, their 28-day and six-month mortality as well as their duration of hospital stay was not significantly different to that of the patients who did not receive CVVH. Hence, it can be claimed that CVVH did have a positive impact on the clinical outcome of the patients who received it as part of the treatment of septic shock.
References
Abstract 0042 – E-Poster EPM.142
Procalcitonin clearance as a prognostic indicator in severe sepsis in critical care
Russells Hall Hospital
Abstract 0047 – E-Poster EPM.143
Validation of the q-SOFA score in sepsis
1Manchester Royal Infirmary
2University of Manchester
The third international consensus definitions for sepsis and septic shock (Sepsis-3) were published in 2016. This document set out recommendations for the definition and diagnosis of sepsis. It was designed to complement the surviving sepsis campaign guidelines (SSC).1,2
Organ dysfunction can be represented by an increase in the sequential organ failure assessment (SOFA). This document introduces the q-SOFA (quick SOFA) score as a diagnostic adjunct in organ failure. This is a validated simple bedside system comprising three criteria; altered mentation, respiratory rate greater than 22 breaths per minute and systolic pressure less than 100 mmHg. It is of use in predicting organ dysfunction and should prompt clinician to initiate further investigation and management.
The aim of our study was to assess the validity of the q-SOFA score in a group of septic patients. We collected data on demographics, the management of sepsis and outcome. We collected q-SOFA, SOFA scores, SIRS criteria and EWS on admission. APACHE 2 scores were calculated on discharge. We compared the q-SOFA to the other scoring systems using Cronbach’s alpha test and used Chi squared test to assess the scoring systems as predictors of outcome.
Forty consecutive patients who had a diagnosis of sepsis were studied prospectively. The 30-day mortality was 22.5%. Thirty-seven patients had an SIRS score of 2 or more. The mean SOFA score was 6.2 and the mean APACHE2 score was 19.
We applied Cronbach’s alpha test to assess agreement between tests. Agreement between SIRS and qSOFA was poor (α = 0.57), although it was superior to comparisons between qSOFA and SOFA (α = 0.28) and SIRS compared to SOFA (α = 0.06).
We applied Chi squared test to the data to see if the scoring data allowed us to assess a significant association with outcome. None of the scoring systems showed a significant association with 30-day mortality. It was most accurate when the SOFA score was used and when the score exceeded 6 points the p-value was 0.12. The p-value for q-SOFA was 0.42 and that for SIRS was 0.26.
We conclude that q-SOFA is not an accurate predictor of 30-day mortality but its superior agreement to SIRS than standard SOFA may confer an advantage. In addition, its simplicity and ease of use at the bedside would be a useful adjunct in the assessment of sepsis and organ dysfunction.
References
Abstract 0104 – E-Poster EPM.145
Widening the indications of non-invasive ventilation in intensive care units: The experience of a district general hospital
1Warwick Medical School
2Warwick Hospital
Indications for non-invasive ventilation of the 47 patients receiving NIV during 2015.
T1RF: type 1 respiratory failure; T2RF: type 2 respiratory failure; BiPAP: two level positive airway pressure; PEEP: positive end-expiratory pressure; EBM: evidence-based medicine.
Outcomes of non-invasive ventilation in 47 patients, grouped by indication. Outcome was assessed as overall 90-day mortality.
Lobar consolidation – The impact of lobar consolidation on NIV outcomes is controversial: the utility of NIV for a diganosis of pneumonia is debated, whilst in patients with COPD complicated by lobar consolidation NIV outcomes are known to be worse.3 In this series of patients receiving NIV, 20 had documented lobar consolidation. These patients had a worse 90-day survival (45%) than COPD patients without consolidation (67%), although with small sample sizes, this was not significant. Examination of prognostic variables amongst patients with lobar consolidation showed that patients with elevated leukocytes were less likely to have a good NIV outcome (OR = 0.827, p = 0.043).
References
Abstract 0082 – E-Poster EPM.146
A service evaluation of the respiratory support available to adult in-patients with cancer in hypoxic respiratory failure
Barts Health NHS Trust
There are a number of treatment options for hypoxic respiratory failure (HRF), and there is an increasing establishment of Optiflow™ (OF) in many Trusts nationwide. High-flow nasal cannula (HFNC) oxygen therapy, delivered via OF, is increasingly used to improve oxygenation, and treat HRF due to its ease of implementation, tolerance and clinical effectiveness. However, to date, no research has been published investigating the use of this adjunct in the cancer population in the ward setting, although much is written within the critical care environment. Indeed, Spoletini et al.1 have recommended that OF use be limited to intensive care units (ICUs).
Using a service evaluation design, and quantitative descriptive statistics, outcomes were evaluated for this patient group during a three-month period. Adult cancer patients within the ward environment were eligible for inclusion when their supplemental oxygen requirements hit an inspired fraction of 0.4 or over. Outcomes collected were: physiological observations (including the National Early Warning Score (NEWS)); the use of OF, the need for transfer to the ICU, how many patients trialled non-invasive ventilation (NIV), how many patients were intubated, and how many patients died. The results of this service evaluation demonstrated that all adult cancer patients in HRF went onto HFNC on the ward, seven (50%) patients were admitted to ICU, six (43%) patients were ventilated, and five (36%) patients died. There was no statistically significant difference found between the patients RR or PF ratio pre and post implementation of HFNC. Practice has seemed to have changed for patients in HRF, as no patients went onto NIV during this period.
The correlation between being ventilated and death was statistically significant (p = 0.002). The median NEWS score was not affected by the instigation of OF, and remained high at an aggregate score of 10 both pre and post OF set up.
In conclusion, the addition of optiflow at ward level does not decrease the acuity of patients in HRF, and the average day of admission to ICU was five, which is longer than current evidence base recommendations. Patients were left too long on the ward in HRF without other options for the treatment of HRF being explored. Further work is needed in this area, particularly at ward level.
Reference
Abstract 0174 – E-Poster EPM.147
The use of electronic patient observations tracking and targeted critical care outreach to reduce in-hospital cardiac arrests
East Sussex Hospitals NHS Trust
An electronic observations system (VitalPAC) was introduced in both hospitals of an acute NHS trust in March 2014 and was fully functional by September 2014. A small, dedicated team implemented training and its use; there was good engagement and acceptance of the technology. The aims were to focus on basic observations on wards, increase the percentage of observations being taken on time, and facilitate earlier alerting to medical teams when patient deterioration occurred, thus allowing earlier intervention. Ultimately, the aim was to reduce preventable cardiac arrests which were preceded by deterioration in the previous 24 h.
The VitalPAC project was run jointly between Critical Care, Senior Nursing and Critical Care Outreach teams. Ward performance data for numbers of observations on time were produced monthly. A performance dashboard allowed comparison between wards. Ongoing feedback, education and training were provided to ward staff, and the key aims of early escalation and cardiac arrest prevention reinforced. VitalPAC was also used by outreach teams to identify those patients with NEWS scores of 5 or more in order to review them on the ward proactively. The use of MET or SET calls was actively encouraged for patients with an NEWS of 9 or more. Cardiac arrest data were compiled, including NEWS scores for the preceding 24 h. Cardiac arrests which occurred in patients with a previously deteriorating or high NEWS score were considered preventable.
Review and comparison of cardiac arrest data for the periods of October 2014–March 2015 (soon after introduction of VitalPAC) and October 2015–March 2016 showed reductions in both the total number of cardiac arrests and also in the proportion of arrests that occurred in patients with a high NEWS score in the preceding 24 h. The number of unpredictable cardiac arrests remained the same.
The use of an electronic observations tracking system led to an increased percentage of observations occurring on time. There was also a significant change in practice of outreach teams, with a much more proactive approach in seeking out deteriorating patients and encouraging escalation plans by ward teams. The overall effect has been a reduction in cardiac arrests, particularly those which were potentially preventable.
Cardiac arrests per 1000 bed days. Cardiac arrest numbers.
Abstract 0225: E-Poster EPM.148
The Lactate 4: Improving the recognition and escalation of deteriorating patients
West Suffolk Hospital
Elevated blood lactate is a common feature of critical illnesses; the magnitude of lactate rise corresponds to severity of illness and consequently to increasingly poorer outcomes.1 The specific cause(s) of such a disturbance varies, but may be generalised to either increased lactate production and/or decreased lactate clearance. The ‘clearance’ of raised lactate (more correctly normalisation) is both a therapeutic target and a marker for the efficacy of treatment of the underlying cause.2,3 Assessment of lactate clearance has recently gained favour as a surrogate measure of circulatory adequacy, and in so doing has begun to displace other metrics such as central venous oxygen saturation (ScvO2).2,3
Our institution has established a Deteriorating Patient Group (DPG) with multi-disciplinary representation from Critical Care, Education & Outreach and Resuscitation Services, and other acute areas. The DPG is both reactive and proactive in providing a medium for organisational learning from preventable adverse outcomes and implementing change to prevent such instances in the future.
When reviewing episodes of preventable or missed/slowly-recognised deterioration, some recurring themes emerge. One such theme is the failure to either measure blood lactate levels in deteriorating patients, or the failure to respond to elevated lactates including the failure to repeat lactate to help assess adequacy of treatment. These instances are commonly associated with a delayed recognition of severity illness, delayed escalation to critical care and possibly poorer outcome(s).
In response to these instances, the DPG has designed and implemented a ‘Lactate 4’ campaign, the key features of which are (1) measuring lactate in unwell patients, (2) considering cause(s) of raised lactate and commencing appropriate treatment, (3) ensuring abnormal lactates are repeated and (4) ensuring appropriate escalation. The Lactate 4 campaign is designed to co-ordinate with local sepsis guidelines and ‘Sepsis 6’ checklists, and hospital escalation policies. Laminated prompts are placed at highly visible locations including communal blood-gas analysis machines.
The Critical Care Education and outreach team are the primary group driving the adoption of this process amongst clinical staff; once alerted, this team assist with ward-based interventions and facilitate expedited referral for Critical Care admission as appropriate.
Lactate 4 aide-memoire.
References
Abstract 0208 – E-Poster EPM.149
Referrals to critical care: Are we reviewing and escalating patients appropriately?
AS Tabish1 and
1Royal Wolverhampton NHS Trust
2Walsall Manor Hospital
Reference
Abstract 0049 – E-Poster EPM.150
Introducing the Bazett formula into the MAP equation to account for varying heart rates: A physiological approach
1Imperial College London
2Royal London Hospital
3East Kent University Hospital Foundation Trust
Validation of the various MAP formulae generated.
Comparison of MAP formulae performance when validated with MSR and AIC. Bazett corrected MAP has the lowest MSR and AIC.
Comparison of the standard and the proposed MAP formulae against the measured MAP.
References
Abstract 0100 – E-Poster EPM.151
Defining the temporal pattern and clinical correlates of lesion expansion using computed tomography following traumatic brain injury
E Carroll1, J Coles1,
1Division of Anaesthesia, University of Cambridge
2Department of Computing, Imperial College London
Lesion progression is a common finding in traumatic brain injury (TBI) when imaging is performed within a few hours of ictus, and is associated with poor outcome.1–3 While previous studies have concentrated on defining evidence of haemorrhagic transformation within cerebral contusions,1–3 the aim of this study was to differentiate between haemorrhage and oedema using serial computed tomography (CT).
Data were collected from patients admitted to a Neurosciences Critical Care Unit with TBI and visible lesions on CT imaging. Using a semi-automated tool specifically designed for segmenting brain (Imseg) and manual delineation, lesions were defined as haemorrhagic core, oedema, traumatic subarachnoid haemorrhage, subdural and extradural haematomas (Figure 1). Imaging findings were compared with clinical management, intracranial pressure (ICP) and patient outcome using the extended Glasgow Outcome Score (eGOS).
The mean (range) haemorrhagic core volume was 9.8 (0.1–49.1) ml on baseline CT (<6 h) and expanded by an average of 78% to 17.4 (0.2–77.7) ml on repeat imaging obtained between 6 and 24 h post injury. Eight patients underwent surgery for evacuation of an EDH, SDH or haemorrhagic lesion within 48 h of injury. While there was little change in the volume of haemorrhagic core after 6–24 h, the volume of pericontusional oedema continued to increase until 7–10 days post injury (Figure 2). Patients with a clinically significant increase in oedema (>25 ml) tended to suffer more episodes of physiological instability within the first 72 h, and needed increased medical therapy for ICP control with two patients requiring decompressive craniectomy on day 7 post injury. There was a non-significant trend for increased ICU stay and poor outcome (eGOS) with the volume of oedema at 7–10 days post TBI (Spearman Rank rho = 0.3, p = 0.14 and rho = −0.3, p = 0.16, respectively).
Initial findings emphasise that patients who undergo imaging within 6 h of injury should have a repeat CT at 24 h since haemorrhagic lesions may expand and require surgical evacuation. Lesion expansion after this time typically relates to worsening oedema and highlights the need to optimise cerebral perfusion and oxygenation in TBI patients and limit any increase in ICP associated with worsening oedema.
Lesion progression following TBI. Haemorrhagic core (green), pericontusional oedema (light blue) and SDH (dark blue). (a) Baseline CT. (b) Repeat CT at 74 h. (c) Repeat CT at 122 h. Contusion progression.
References
Abstract 0125 – E-Poster EPM.152
The association of intravenous fluid administration on patient outcomes in critical care
C Gerrard1,
1Papworth Hospital
2University of Cambridge
p < 0.05.
References
Abstract 0181 – E-Poster EPM.153
Fluid management following the resuscitation phase of critical illness: A survey of intensive care specialists
E Fan1, A Ferguson2, J Marshall1,
1University of Toronto
2Southern Health and Social Care Trust
3Queen's University of Belfast
4Regional Intensive Care Unit, Royal Victoria Hospital
The accumulation of large volumes of fluid is common in the critically ill and is associated with adverse clinical outcomes including mortality.1 Strategies aimed at prevention or treatment (deresuscitation) of fluid overload may be beneficial following haemodynamic stabilisation.2
To inform the design of future trials on this topic, and to better understand current practice, we invited specialist (consultant) members of the Intensive Care Society (n = 1550) and European Society of Intensive Care Medicine (n ≈ 11,500) to complete an online survey. The survey consisted of demographic, attitudinal and practice-based questions, together with case vignettes and associated therapeutic options, and utilised a combination of Likert scales, multiple choice, and free text responses.
A total of 526 responses were received. The majority of respondents (n = 289 of 339, 85%) perceived fluid overload as a modifiable source of morbidity resulting from fluid administration, endocrine influences and acute kidney injury. A total of 379 of 432 respondents (88%) regarded the topic as an important one for future research and 87% (n = 374 of 432) expressed willingness to enrol patients in a randomised trial comparing deresuscitation with usual care.
Example case vignette.
Our survey identifies the perceived importance of fluid overload as a modifiable determinant of morbidity in critically ill patients. Wide variability in clinical practice, and uncertainty as to the risks and benefits of deresuscitative strategies highlight the need for robust clinical trials of alternative fluid strategies in critically ill patients.
References
Abstract 0181 – E-Poster EPM.153
Fluid management following the resuscitation phase of critical illness: A survey of intensive care specialists
E Fan1, A Ferguson2, J Marshall1,
1 University of Toronto
2 Southern Health and Social Care Trust
3 Queen's University of Belfast
4 Regional Intensive Care Unit, Royal Victoria Hospital
The accumulation of large volumes of fluid is common in the critically ill and is associated with adverse clinical outcomes including mortality.1 Strategies aimed at prevention or treatment (deresuscitation) of fluid overload may be beneficial following haemodynamic stabilisation.2
To inform the design of future trials on this topic, and to better understand current practice, we invited specialist (consultant) members of the Intensive Care Society (n = 1550) and European Society of Intensive Care Medicine (n≈11,500) to complete an online survey. The survey consisted of demographic, attitudinal and practice-based questions, together with case vignettes and associated therapeutic options, and utilised a combination of Likert scales, multiple choice, and free text responses.
A total of 526 responses were received. The majority of respondents (n = 289 of 339, 85%) perceived fluid overload as a modifiable source of morbidity resulting from fluid administration, endocrine influences and acute kidney injury. A total of 379 of 432 respondents (88%) regarded the topic as an important one for future research and 87% (n = 374 of 432) expressed willingness to enrol patients in a randomised trial comparing deresuscitation with usual care.
Responses to a sample case vignette are shown in Table 1. Most respondents (n = 355 of 376, 94%) reported that fluid overload was a common occurrence and used diuretics daily (n = 98 of 376, 26%) or frequently (n = 152 of 376, 40%) in ICU practice. Most respondents favoured loop diuretics alone either by intermittent bolus dosing (n = 202 of 372, 55%) or by infusion (n = 43 of 372, 12%) as the initial approach to deresuscitation. Reported use of adjunctive hyperoncotic albumin infusions and non-loop diuretics use was infrequent. The majority of respondents were willing to initiate diuretics despite a noradrenaline dose of up to 0.1 µg/kg/min, and to continue despite complications such as mild hypotension (mean arterial pressure: 55–65 mmHg) or dysnatraemias (serum sodium: 132–135 or 145–150 mmol/l).
Our survey identifies the perceived importance of fluid overload as a modifiable determinant of morbidity in critically ill patients. Wide variability in clinical practice, and uncertainty as to the risks and benefits of deresuscitative strategies highlight the need for robust clinical trials of alternative fluid strategies in critically ill patients.
References
Abstract 0248 – E-Poster EPM.154
Resuscitative endovascular balloon occlusion of the aorta: A systematic review of the literature
E Dedola and
Queen Alexandra Hospital
Trauma remains a leading cause of morbidity and mortality globally. Deaths associated with traumatic haemorrhage continue to be seen. Evidence suggests that deaths attributable to haemorrhage associated with certain characteristics, particularly non-compressible torso haemorrhage (NCTH), may be preventable.1 Furthermore, methods of haemorrhage control, prior to surgical intervention, are limited. These considerations are of paramount importance to critical care. A novel method of haemorrhage control for NCTH, REBOA (Resuscitative Endovascular Balloon Occlusion of the Aorta) has emerged as a potential haemorrhage control device in this context, although it is associated with recognised concerns.2
The authors undertook a systematic review to assess the current evidence in the medical literature regarding the use of REBOA in non-compressible torso haemorrhage.
An English language literature search was performed through Embase and Medline via the NICE evidence forum using the search terms (‘Trauma’ OR ‘Shock’ OR ‘Haemorrhage’) AND ‘Resuscitative Endovascular Balloon Occlusion of the Aorta’/‘REBOA’. Identified studies subsequently underwent protocolised screening to assess eligibility for review. Furthermore, assessment was made of the reference lists of eligible studies, with papers of interest not previously identified by the literature search process subjected to the same strict entry criteria prior to review. Only papers identified via these criteria were included in subsequent data analysis. The study period for inclusion in this systematic review was 1 January 2000 until 30 June 2015.
We identified 34 papers for primary screening. Following initial exclusions, 31 papers entered the formal screening process, of which 17 were eligible to enter qualitative data synthesis. No meta-analyses were produced on these heterogeneous studies.
Results demonstrated current use of REBOA in global trauma systems. We identified significant differences between studies. These included: indications for use; REBOA Operator clinical background/training; catheter placement technique; final placement position within Aorta; outcomes (physiological, morbidity and mortality) and complications. Reported complications were noted to be potentially significant. Several studies highlighted educational programmes and technical feasibility, particularly in the context of pelvic trauma. However, despite obvious promise, we conclude that the role of REBOA in haemorrhage control in trauma is yet to be fully defined. Further research is required to enable conclusions to be drawn regarding the appropriate use of REBOA in this setting. It is essential that, given increasing interest in (and potentially increasing usage of) REBOA, critical care practitioners have an understanding of current practices, and potential complications, related to this innovative technique.
References
Abstract 0056 – E-Poster EPM.155
Factors contributing to post-operative AKI in patients undergoing major bowel surgery
S Abdul, W Butt, M Chablani, S Panjwani, Y Plesnikova, M Ridley, M Suleiman,
Pilgrim Hospital Boston
Abstract 0033 – E-Poster EPM.156
Changing renal replacement fluid to improve phosphate control
NHS Grampian
Phosphorus is a key element in many biological processes including cell membrane synthesis, glycolysis, ATP and 2,3-DPG synthesis.1 Furthermore, hypophosphataemia is common in the ICU, especially on renal replacement therapy (RRT).2 A previous study in our ICU found a 25% incidence of hypophosphataemia in patients requiring RRT. This study looked at the change in this with the use of a phosphate-containing replacement fluid during continuous RRT.
Our standard first-line replacement fluid was changed from Prismasol (Gambro, 0 mmol/l phosphate) to Phoxilium (Gambro, 1.2 mmol/l phosphate). Our protocol for RRT delivery was otherwise unchanged. While on RRT, all patients had their phosphate levels reviewed daily. The normal range for phosphate was taken as 0.8–1.5 mmol/l.
A total of 23 patients were reviewed, with a total of 109 RRT-days. Patients were on RRT for a median of three days (min 1, max 15). There was a 1.83% incidence of hypophosphataemia (2 RRT days). The incidence of hyperphosphataemia was 62.39% (68 RRT days) overall, dropping to 49.54% (54 RRT days) when those with pre-existing hyperphosphataemia were excluded. Of these cases, 92.59% (50) were mild (1.6–2.1 mmol/l) and 7.41% (4) were severe (>2.1 mmol/l).
This study confirms that using a replacement solution containing phosphate is effective in reducing the incidence of hypophosphataemia in our ICU patients requiring continuous RRT. We did see an increased incidence in hyperphosphatemia; however, this was not associated with any patient harm and, unlike hypophosphataemia, there is little in the literature to describe how it effects outcomes in the ICU population. Approximately 50% of the hyperphosphataemic days were in patients on VV-ECMO. These patients have other reasons to have a raised phosphate load (RBC transfusion, pump haemolysis) which may have been exaggerated by the use of phosphate-containing replacement fluid.
Phoxilium is now the standard first-line replacement fluid for patients on continuous RRT in our department. Further work should be done locally to see if a different protocol is required for patients on VV-ECMO.
References
Abstract 0160 – E-Poster EPM.157
Haemodiafiltration using regional citrate anticoagulation: An effective and cost saving strategy. A single centre experience
N MacCallum,
University College London Hospital
In recent years, anticoagulation strategies have been adapted to improve the safety and efficacy of renal replacement therapy in critically ill patients. Pre-filter regional citrate anticoagulation has been compared with heparin in several clinical trials, and some have suggested that citrate was superior in bleeding risk and efficiency but no significant difference was found in patient or renal outcomes.1 There is very limited evidence in the literature on the cost effectiveness of citrate-haemofiltration and haemodiafiltration. Shilder et al.2 reported lower cost with citrate-haemofiltration due to cheaper cartridge and reduced labour with longer lasting filter sets. However, another study has found citrate-haemofiltration more expensive compared with heparin, because of its associated costs.
We reviewed our early experience in haemodiafiltration with regional citrate anticoagulation vs. standard practice with heparin-haemofiltration and compare their cost effectiveness in 2015–2016.
Forty-three patients with heparin anticoagulation and Eighty-five patients with citrate anticoagulation were included in this analysis. Twenty-one per cent less patients in the citrate group required blood transfusion, saving on average one unit of packed red cells and £40 per patient filter day. There was no significant difference in the risk of haemorrhage between the two groups. The average number of coagulation test performed per patient was 14 in heparin group vs. 7 in citrate group, saving a mean of £180 per patient. Large volume of dialysate and replacement fluids, up to 4000 ml/h has been reported in the published literature, may be required. This contributes to a large part of the haemodiafiltration cost. At our centre, a lower flow rate strategy is used for all patients (2200 ml/h). Mean filter life in the citrate group being 44 h (vs. 32 h in the heparin group), which is comparable to that in the reported literature (34–46 h).1 The difference in filter life span between the two anticoagulation groups is also comparable to the reported literature (15 h).1 20% and 24% of filter cartridges clotted in the citrate and heparin group respectively, which is both lower than that reported in the literature.2
We conclude that, in our experience, lower flow rate haemodiafiltration with regional citrate anticoagulation is a safe and cost effective method of renal replacement therapy. Further study is recommended to evaluate the impact of lower flow rate strategy on patient and renal outcomes.
References
Abstract 0045 – E-Poster EPM.158
Practice and outcomes of renal replacement therapy within Aberdeen Royal Infirmary Intensive Care Unit
Aberdeen Royal Infirmary
Data were available for continuous RRT in 57 patients. Nine patients (16%) were prescribed appropriate doses of 20–30 ml/kg/h. Further 48 patients were prescribed high-dose RRT up to 120 ml/kg/h. The mean dose was 51 ml/kg/h. The average length of treatment was 51 h. Twenty-two patients (39%) had treatment for less than 24 h and only four of these survived. The remaining 25 patients either had intermittent haemodialysis or we were not able to review their notes.
References
Abstract 0111 – E-Poster EPM.159
Improving continuous end tidal carbon dioxide monitoring in critical care
S Amdekar, A Bellini, K Elliot,
Queens Hospital
In an effort to increase the rate of monitoring, the audit data were presented at the departmental teaching for all members of the Critical Care team. This involved doctors and allied health professionals including physiotherapists and dieticians. The audit results and rational behind continuous ETCO2 monitoring were discussed, and all health professionals were encouraged to ensure monitoring in their patients. In this forum, the barriers to effective monitoring and ways to remove them were also discussed.
The Critical Care Matron designed teaching with nursing team leaders. They taught their teams about the practicalities of ETCO2 monitoring, its rationale and its importance via individual bedside teaching and enforced the monitoring of ETCO2 on their shifts. In May 2016, a new ventilator system was introduced which has integrated CO2 monitoring, providing a sampling line is attached. Prior to this, monitoring was provided via a limited number of external modules which were attached to the patients’ monitor and anecdotally were unreliable; often spontaneously breaking requiring multiple changes during a shift.
References
Abstract 0112 – E-Poster EPM.160
Who does your tracheostomies and when?
St Helier Hospital
The TracMan trial demonstrated no significant benefit to early tracheostomy (<10 days).2 It also highlighted that we are poor at predicting patients needing prolonged ventilation from the outset. Our impression is that most intensivists’ practice is consistent with the study’s findings with most electing to wait until >10 days of mechanical ventilation before performing tracheostomy. Broadly speaking, our practice is in line with this with 59% performed after >10 days. Interestingly, those patients who underwent an early tracheostomy had a higher mortality.
References
Abstract 0163 – E-Poster EPM.161
Audit of current central venous catheter (CVC) insertion depth
Dumfries and Galloway Royal Infirmary
Correct central venous catheter (CVC) placement is not universally agreed; however, there is some consensus that placement in the lower superior vena cava or upper right atrium may help avoid certain complications.1 The most common method of identifying CVC placement is by plain chest radiograph, but this is limited by the 2D nature of plain films and anatomical variations. However, there is research showing reliable anatomical landmarks for satisfactory CVC placement.2,3 Subsequently, we devised a visual tool to aid identifying correct placement on plain films and retroactively applied this tool to patients admitted to our ICU over a 15-month period.
Wardwatcher was used to identify the patients, and then chest radiographs were inspected. We identified 200 patients who had had a CVC during their ICU stay. One Hundred and thirty-one patients (65.5%) had satisfactory CVC tip position according to our criteria. Nine (4.5%) were identified as being placed too long that they might be beyond the pericardial inflection. Fifty-nine (29.5%) were identified as being too short. Three of the lines classed as too short ended at an acute angle to the trachea-risking vessel wall abutment.
Visual aid for identifying correct line tip position.
One patient was excluded as they had an internal jugular CVC that diverted in to the subclavian vein. They did not have a subsequent radiograph with a line in situ.
In our population, catheters that were not inserted far enough outnumbered those considered too long. Short insertion increases risk of thombosis or incorrect placement in the braciocephalic or azygous veins. This can also effect CVC function with poorer flow rates attainable for renal replacement and more venous irritation from hyperosmolar fluids/drugs that are not as rapidly diluted. We argue that it is more common to be too short but be unable to further advance the CVC thus requiring new insertion wasting time and resources. Conversely, if a catheter is identified as too long it can be pulled back and resecured.
We plan to now perform a prospective audit looking at correct tip placement to try and identify an appropriate line insertion length based on patient gender, height and weight. Doing so would hopefully reduce the number of incorrect CVC placement and minimise risk of complications.
References
Abstract 0078 – E-Poster EPM.162
Percutaneous tracheostomy safety on the intensive care unit… where can we improve?
M Chopra and
Plymouth Hospitals NHS Trust
Tracheostomies are common procedures in the intensive care environment (ICU), occurring in a challenging cohort of patients. According to a recent analysis, percutaneous tracheostomy-related deaths occur in 1 in 600 procedures with the majority of complications occurring in the ICU.1,2 Without the necessary competencies to manage daily tracheostomy care and potential complications arising, patient safety can be compromised. Identifying key areas of concern is paramount to allow targeted simulation-based training packages to be developed in response to need. Following the design and implementation of a WHO style percutaneous tracheostomy insertion safety checklist, self-reported staff competence and confidence in both planned and emergency situations was assessed utilising a non-identifying, web-based questionnaire of criteria outlined in the National Tracheostomy Patient Safety (NTSP) tracheostomy care competency statement. A 30% (n = 65) response rate was achieved. Self-reported competence levels were reassuring with regard to tracheostomy indications (100%); assistance with tracheostomy insertion and complications arising (89%); post insertion respiratory assessment and tracheostomy-specific checks (85%); post insertion documentation and assistance with decannulation (86%). Key areas of concern identified related to the description of life-threatening red flag indicators (64%) and familiarity with emergency management algorithms for tracheostomy (77%) and laryngectomy (72%) patients. The percutaneous tracheostomy insertion safety checklist has been effective in improving baseline staff competence relating to peri-insertion procedures and future simulation training packages across the staff disciplines need to focus on the relatively low frequency, high stakes emergency situations with follow-up assessment of competence and confidence following their implementation.
References
Abstract 0063 – E-Poster EPM.163
Prone to inconsistency? What is the usage of prone ventilation amongst acute respiratory distress syndrome patients at a large district general hospital?
S Edwards, M Holland,
Royal Cornwall Hospital
Acute respiratory distress syndrome (ARDS) is a syndrome of non-cardiogenic, inflammatory pulmonary oedema which may lead to catastrophic refractory hypoxia. Prone ventilation is postulated to improve this through a variety of mechanisms including optimisation of regional ventilation-perfusion differences and shunt. However, there is challenging decision making regarding patient selection, timing and practical feasibility on the ICU at the time. This project aimed to evaluate the proning strategy employed as a unit for our patients with ARDS, and assess whether we are proning patients who the literature says will benefit.
Single-centre retrospective audit of patients diagnosed as ARDS. The unit database was screened for adult patients ventilated for greater than three days between January 2015 and January 2016. Notes and chest imaging were then reviewed to check for bilateral infiltrates consistent with ARDS and to exclude cardiac failure. Patient characteristics, classification into mild/moderate/severe by PaO2:FiO2 ratio, 6-h ventilatory data, underlying diagnosis, 30-day mortality, and whether the patient was proned or not were recorded. These data were compared against criteria for proning in the Proseva trial1 (PaO2:FiO2 <150 mmHg, FiO2>0.6, PEEP>6 cmH2O, tidal volume 6 ml/kg), and for use as rescue therapy (PaO2:FiO2<100 mmHg).
A total of 176 patients were ventilated for three days or more over the study period, and 135 were excluded after review of notes and imaging. Therefore, 41 patients were included as having ARDS. Median age was 58. Median APACHE-II score was 20. Eighty per cent had primary lung insult, e.g. pneumonia. Six were classified as mild, 32 as moderate, 3 as severe based on median P:F ratio. Overall 30-day mortality was 17%. Seven patients were proned – one in the mild group, four moderate, two severe. Two out of seven proned patients died within 30 days. Twenty-seven patients were eligible for proning by Proseva criteria, of which seven of these were proned. Twenty-four patients met criteria as rescue therapy, of which six were proned.
Proning is potentially a powerful tool in the management of refractory hypoxia in ARDS. The patients who will benefit from proning remain debated, but comparison of our hospital data to inclusion criteria for positive trials has revealed heterogeneity in our proning strategy. It is possible that the data to explain this fell within the 6 h gaps in ventilatory data, or practical considerations prevented proning. However, given debate over evidence for proning, deviance from the literature base is perhaps unsurprising. The introduction of a Standard Operating Procedure for Proning should reduce variance and improve performance.
Reference
Abstract 0119 – E-Poster EPM.164
HIV testing in critical care: An audit towards opt out testing
H Buckley1, A Donnelly1,
1Bolton NHS Foundation Trust
2Salford Royal NHS Foundation Trust
The number of people living in the United Kingdom with HIV has been rising, with a quarter of those undiagnosed.1 Individuals with HIV who present late with low CD4 counts have higher morbidity and mortality, and early diagnosis has marked impact on health related costs and outcomes.2 The national UK guidelines for diseases and presentations where HIV infection should be tested3 were used as a benchmark for assessing the departments current testing process.
A case note review of all admissions to a district general critical care unit from 1 January 2015 to 31 December 2015 was undertaken. Diagnoses consistent with the recommended UK national guidelines were sought.3 HIV tests conducted by the trust in the time period were interrogated and those whose location was in critical care were cross-referenced with the case notes.
There were 382 patients admitted to critical care in the timeframe scrutinised, with six re-admissions and no notes excluded. No patients were known to have a HIV diagnosis prior to admission or had tested positive prior to ICU admission. There was no HIV testing protocol and clinician judgement was used in deciding whom to test. Twenty-eight patients were tested for HIV, and all had presentations which were recommended for testing by the 2008 guidelines.3
Had the 2008 guidelines been followed, then 96 other patients would have been tested. Six of the 28 patients tested returned positive results, giving a crude prevalence of 15.7 per 1000 of critical care admissions per year. Four of the six patients with HIV had advanced disease and did not survive to ICU discharge.
Our ICU prevalence of 15 per 1000 patients was eight times higher the reported 1.8 per 1000 local prevalence. Options for improving testing rates were looked at and an opt out model is being implemented. Careful consideration about capacity, consent and positive test disclosure and follow-up has been made in discussion with the Sexual Health team and the trust legal department. A prospective study will be undertaken.
References
Abstract 0038 – E-Poster EPM.165
Fluid balance and electrolytes in aneurysmal subarachnoid haemorrhage – Non-prescribed fluids as a cause of hypervolaemia?
Nottingham University Hospitals NHS Trust
Aneurysmal subarachnoid haemorrhage (SAH) carries a high burden of mortality and morbidity. While particularly in the context of vasospasm, triple-H therapy (hypertension, haemodilution and hypervolaemia) is still practised, recent studies suggest that this strategy is not effective and that hypervolaemia may be detrimental. We audited our fluid and electrolyte management of critical care patients with aneurysmal SAH.
The records of 41 consecutive patients admitted to critical care with aneurysmal SAH (between October 2014 and April 2015) were reviewed.
Seventeen (41%) of patients were male. The mean age on admission was 57.7±9.8 years. The median volume of prescribed fluid was 2405 ml (1250–3200 (0–5465)). The median total fluid volume received was 3645 ml (2475–4965 (1220–7697)). Median fluid balance during the first 24 h in critical care was 1040 ml (36–2375 (−1500–5390)).
The mean admission sodium was 138 ±5mmol/L and did not vary with gender. In 21 (51%) patients, there was a ≥5mmol/l change in serum sodium concentrations (of which 5 (12%) had >10 mmol/l change). Greater changes in sodium occurred with a higher WFNS grade.
Eight patients (20%) had hyponatraemia on admission, and five (12%) developed it following admission. Five days later, serum sodium in 26 patients (64%) was within normal limits (135–145mmol/L), five (12%) were hyponatraemic, three (7%) were hypernatraemic. Six patients (17%) were deceased. No association with overall fluid balance was noted.
Many patients had a large positive fluid balance, though serum sodium concentration was not impacted. This may be due to fluid administration being altered to maintain sodium, and prescription of fluids without consideration of other contribution to overall fluid balance. We recommend that fluid administration should be closely tailored to losses in order to maintain euvolaemia, whilst closely monitoring sodium levels.
Abstract 0091 – E-Poster EPM.166
Admissions to ####### ICU of patients with hepatic failure: A retrospective review
A Ashton and
Basingstoke Hospital HHFT
Hepatic failure is a rare, complex, condition.1 Despite advances in critical care, mortality rates remain high.2 Furthermore, this patient group accounts for increasing rates of critical care admission. Several prognostic scoring systems are available to clinicians to guide assessment and management of critical care patients. However, these are not specific to the hepatic failure patient. A recent paper in Critical Care3 validated two scoring systems specific to this population, the Child – Turcotte Pugh and Lactate (CTP + L) Score and the Royal Free Hospital (RFH) Score.
We reviewed admissions of patients with hepatic failure to ####### ICU, retrospectively applied the CTP + L and SOFA Score prognostication systems, reviewed predicted mortality at time of admission and subsequently compared this data to identified outcomes.
Patients identified using diagnostic coding entered into the ICNARC database. Identified episodes underwent data collection on aetiology of admission; physiological parameters at admission; length of stay (ICU and hospital) and outcome. Collated data subsequently processed to calculate CTP+L and SOFA scores, with comparison made with data published from Royal Free London and Glasgow.3 Study period – January 2010 until May 2015 inclusive.
ICNARC database identified 32 episodes of hepatic failure requiring admission to ICU, predominantly males aged 50–70 years. Alcohol excess was a strong determinant of development of hepatic disease, identified in 78% of cases. Commonest causes of hepatic decompensation and subsequent hepatic failure were sepsis (41% of cases) and upper gastrointestinal haemorrhage (31%). Following admission to ICU, the mortality rate was 66%, with a further 50% of admissions discharged from ICU surviving <50 days.
Comparison with published mortality rates for the Royal Free, London and the Glasgow Royal Infirmary,5 demonstrated that CTP+L and SOFA Scores were higher on admission to ######## ICU with a diagnosis of hepatic failure (CTP+L Score of 11 for Glasgow and 13 for London, compared to 15.2 for ########). Mortality rates were higher in the ####### ICU group (Glasgow dataset 30% for ICU and 46% for hospital, Royal Free London dataset with 37% ICU and 46% hospital mortality). This review has enabled departmental and multidisciplinary discussion related to areas of potential improvement in care of this complex patient group. We aim to prospectively collect data on admission to ####### ICU with hepatic failure as well as the development of Departmental Audit Standards, including the calculation of CTP+L and SOFA Scores.
References
Abstract 0025 – E-Poster EPM.167
The relocation and road transfer of intensive care patients to a new hospital in Bristol: Our experiences
Southmead Hospital
In May 2014, North Bristol NHS Trust merged its two existing hospitals – Southmead and Frenchay – into a new, purpose-built building. The project involved the movement of 540 patients, many over a distance of several miles. It was one of the largest single patient transfer operations ever conducted in the United Kingdom. We describe the planning processes and transfer of 24 level two and three patients from two intensive care units into the new hospital. These transfers were performed successfully, without significant incident and under intense scrutiny from the Trust, patients and the media. In this, we also reflect upon our experiences of this process, which may be of benefit to those encountering a similar move in the future.
Abstract 0232 – E-Poster EPM.168
In-hospital and long-term outcome following acute stroke. A single centre experience
D Brealey, N Gauge,
University College London Hospital
Stroke is associated with significant in-hospital mortality and morbidity in survivors. Prognostication is an important determinant for decision for admission to intensive care unit (ICU), withholding and withdrawal of treatment. Nevertheless, predictive scoring systems often include impractical variables and or lack predicting accuracy. Mean in-hospital and one to year year mortality have been suggested as 55% (range: 48–90%) and 68% (range: 59–97%), respectively. As a result of the huge variations in patient outcome and the difficulty to confidently prognosticate after acute stroke, management planning and communication with patient’s family remain a major challenge.
We reviewed the outcomes of acute stroke patients who were admitted to our ICU between December 2012 and 2014. We reviewed the patient’s record and contacted patients and patients’ GPs, one to three years following acute stroke events, to follow-up on their progress, using the Modified Rankin Scoring system.
Although acute stroke is associated with high morbidity and mortality rates, our cohort of patients has overall lower mortality rate, with a larger proportion surviving for >1 year after discharged, compared with the reported literature. Prognostication and decision for ICU admission remain a challenge in all age groups. A structured MDT approach both in hospital and the community is vital to patient’s physical and psychological recovery.
References
Abstract 0075 – E-Poster EPM.169
End of life care in the ICU – An audit of practice and results of a quality improvement program
R Docking, K Fairhurst and
NHS Greater Glasgow & Clyde
End of life care is a daily reality in the ICU. Despite this, providing timely and appropriate palliation is not without difficulty. The often rapid deterioration of critically ill patients, the cultural switch from supportive to comfort care, and most recently the criticism of structured end of life care pathways1 can mean that a ‘good death’ on the ICU is difficult to achieve consistently.
Thereafter, we embarked on a quality improvement process which involved dissemination of the baseline results to the department, introduction of an aide memoire system to prompt timely communication, improvements to administrative support, ease of use adaptations of the electronic documentation and improvements to training in these systems for doctors rotating through the ICU.
The documented care and communication levels were calculated using percentages, and the significance determined using contingency tables and Fisher’s exact test (Table 1).
Following her review into concerns around the care of the dying patient, Baroness Neuberger concluded that ‘preventable problems of communication between clinicians and carers accounted for a substantial part of the unhappiness reported1’. We therefore feel that the significantly improved documentation of discussions with families around EOLC, and the excellent documentation of the decision-making process throughout, is encouraging. There was consistent improvement across the direct patient care quality indicators, including formalisation of DNACPR status and the prescription of anticipatory medicines.
The results demonstrate substantial improvement in the production of timely discharge letters; however, we did not demonstrate a statistically significant difference in immediate communication with the patient’s general practitioner.
We conclude that although there are barriers to the provision of excellent end of life care in the ICU environment, simple interventions can lead to significant improvements and help to embed good practice during the difficult period of a large hospital merger.
Reference
Abstract 0051 – E-Poster EPM.170
Intensive care unit admission refusal – A pilot study
A Akinyemi, T Olusanya,
Department of Anaesthesia, University of Ibadan
The need for intensive care exceeds its availability most times because resources are limited.1 Our objectives were to determine the incidence of admission refusal and factors associated with such in our intensive care unit (ICU).
In this prospective observational study, the following were obtained for patients referred to our 6-bed general ICU over a six-week period: age, gender, date and time of referral, source of referral, reason for referral, whether ICU was full or not full at referral, and modified early warning score (MEWS). Others include whether admitted or not, and if not admitted, reasons for admission refusal. Binomial logistic regression analysis was used to determine predictors of ICU admission refusal.
Patients admitted and those denied admission were 37 (50.7%) and 36 (49.3%) respectively. Following univariate analysis, there were no statistical differences in the age (37.65±18.48 vs. 39.63±26.52) and MEWS (5.76±1.92 vs. 5.94±1.51) of patients in the admitted and not admitted groups respectively (p = 0.71 and p = 0.65). Refusal was highest for sepsis (80%) and respiratory failure (71.4%) groups and lowest for severe head injury (18.2%), no statistically significance difference was found in the MEWS for patients with sepsis (6.60±1.35) and those with severe head injury (6.09±1.41) (p = 0.35). Commonest reason for admission refusal was unavailability of ICU bed (81%). Lack of ICU bed was the only independent predictor of ICU admission refusal (OR: 16.58; 95% CI: 3.65–75.42; p = 0.0001).
Univariate analysis of factors associated with intensive care unit (ICU) admission or refusal.
Reasons for admission and modified early warning score.
References
Abstract 0178 – E-Poster EPM.171
Variable clinician perceptions of futility mandate shared decision making in treatment escalation planning and cardiopulmonary resuscitation
Homerton University Hospital NHS Foundation Trust
Recent legal judgement1 informing ‘Decisions relating to Cardiopulmonary Resuscitation (CPR)’ (Resuscitation Council UK)2 suggests that shared decision making (SDM) needs to take place if there is a ‘realistic chance’ that CPR or escalation of care may be successful. This was previously termed as ‘not futile’. If deemed to be ‘futile’, the patient can be informed that a unilateral medical decision has been made not to offer CPR or another intervention.
However, this depends on two assumptions. First, that clinicians can reliably estimate prognosis, i.e. the percentage chance of a good outcome – which research suggests is problematic.3 Second, that there is reasonable agreement on how low a percentage chance of a good outcome would make proceeding with treatment ‘futile’.
Junior doctors and nurses on wards are familiar with changing treatment goals and CPR decisions as consultants’ rotate. The authors postulated that wide variation in the perception of ‘futility’ could underpin variable clinical decision making.
During a postgraduate education session on TEP (treatment escalation planning) and CPR, we asked a mixed cohort of 25 healthcare professionals (8 consultants, 15 junior doctors, 2 nurses) to indicate what percentage chance of a good neurological and functional outcome they would deem futile in the context of CPR. An anonymous voting pad system was employed.
The results demonstrate a complete lack of consensus on what constitutes ‘futility’. This variation is then overlaid on the known difficulties in predicting prognosis accurately. The results show that even if two different clinicians (or an MDT) were able to accurately predict the same percentage chance of a patient having a ‘reasonable outcome’ post-CPR, one clinician would interpret that chance as being ‘futile’ and another as offering ‘a reasonable chance of success’. The legal standpoint, that it is reasonable to not offer CPR or other intervention based on a ‘medical’ decision of ‘futility’, is therefore untenable.
References
Abstract 0095 – E-Poster EPM.172
Not failing the frail: Should you admit the extreme elderly to intensive care?
1Intensive Care Unit, Broomfield Hospital
2Mid Essex Hospitals NHS Trust
Frailty is the loss of physiological reserves, with vulnerability to adverse outcomes.1 Elderly patients with frailty have been shown to have worse outcomes during critical illness,1 with persistent dependence and disability in survivors.2
There is limited data about outcomes in the critically unwell extreme elderly in the UK and the relationship to frailty. We proposed to investigate the relationship between frailty and outcomes in the extreme elderly patient requiring critical care.
All elderly patients (age 80 years and above) admitted to a mixed intensive care unit (ICU) at Broomfield Hospital, between January 2014 and December 2015 were included in the study. Manually curetted data used for ICNARC submission was extracted from the electronic database systems. Frailty was classified using the ICNARC four-point scale: 1 – no assistance with activities of daily living, 2 – minor assistance, 3 – major assistance and 4 – total dependency. ICU and hospital mortality were the primary outcomes.
Standard statistical descriptors were used to summarise data. Generalised linear models with Gaussian or binomial families were used to model continuous and dichotomous outcome variables. Decision trees were constructed to identify variables related to outcome.
Eighteen per cent (n = 305) of all patients admitted to ICU over the two-year study period were aged 80 years and above. The ICU and hospital mortality rates in the elderly were 18% and 24.9%, respectively (13.4 and 17.4% for all age groups, p = 0.15 and p = 0.02).
Mortality increased with the number of organs supported. Advanced respiratory support but not cardiovascular support was associated with mortality. The number of admissions decreased with increasing frailty. However, frailty was not associated with mortality and duration of stay either in ICU or hospital.
In the extreme elderly, APACHE II scores show increasing differences between survivors and non-survivors as frailty increases. In contrast, ICNARC scores are less effective in differentiating non-survivors in patients with high frailty (Figure 1). A four-parameter decision tree could classify ICU survivors (Figure 2).
In contrast to previous studies, our data do not show a correlation between frailty and mortality or duration of stay in the extreme elderly. The magnitude of physiological disturbance is more predictive of mortality than frailty scores. Respiratory but not cardiovascular support is associated with poor outcomes.
These findings imply that even in the extreme elderly good outcomes can be achieved with appropriate patient selection. However, standard predictive modelling is of limited utility in patient selection.
Characteristics of APACHE II and ICNARC scores, in differentiating survivors across different frailty groups. While the difference in APACHE II scores increases as frailty increases, the differences in ICNARC scores decreases. The ICNARC score incorporates frailty in the predictive model. Decision tree to identify probability of death in ICU in patients aged 80 years or above. All parameters are the highest value in the first 24 h of admission. Core Tem = Core Temperature. The value in the ellipse is the highest probability in the class and the two numbers below the label are the probabilities of survival and death respectively
References
Abstract 0144 – E-Poster EPM.173
The frail in intensive care: “Live or let die”?
Intensive Care Unit, Broomfield Hospital
Futility describes treatments to patients that fail to accomplish the intended physiological goals.1 Considering the vulnerability of frail patients, the selection of the ones with least physiological embarrassment should take place in order to benefit the patient. We proposed to investigate the characteristics of the frail elderly patients admitted to the ICU.
We analysed all elderly patients aged 80 and above admitted to a mixed ICU at Broomfield Hospital from February 2015 to February 2016. Frailty was classified according to the Rockwood clinical frailty scale.2 Patients with a score of 5–8 were included in the study.
The data were extracted from electronic records for ICNARC data submission and analyzed via Excel.
Thirty-nine patients out of 150 admitted to the ICU over the one year study were aged 80 and above with a frailty score of 5–8. Twenty-three patients out of 39 (59%) died, 11 in ICU. Four of the non-survivors died 13 days–292 days after hospital discharge.
All of the 11 ICU non-survivors had limitations in care, four of which had active withdrawal of care when treatment futility was identified. The presenting conditions for the ICU non-survivors were: cardiogenic shock, respiratory failure, intestinal perforation, sepsis, cardiac arrest, sepsis post operatively and emergency abdominal aortic aneurysm repair. Their median predicted mortality according to ICNARC was 21 and APACHE score 10.
Twenty-eight patients (72%) survived ICU admission; 13 had limitations in care on admission and on discharge. Eighteen patients had undergone recent surgery. The conditions for the hospital survivors were; revision of total hip replacement, cholecystectomy, renal failure, temporary pacing wire insertion, perforated abdominal viscera, intestinal obstruction and sepsis. Their median predicted mortality according to ICNARC was 20 and APACHE score 10.
There was almost no difference in the predicted mortality according to ICNARC and APACHE scores for the survivors and non-survivors.
A common feature of the survivor group was recently having undergone surgery, demonstrating adequacy of physiological reserve and perhaps contributing to their survival following critical illness. Recent surgery is not adequately weighted for in either scoring system.
Forty-one per cent of frail elderly patients admitted to the ICU survived their critical illness to hospital discharge. Age and frailty score should not be factors that prevent ICU admission.
References
Abstract 0099 – E-Poster EPM.174
An audit of delivered versus prescribed enteral nutrition in adult intensive care
Manchester Royal Infirmary
The delivery of early nutrition can reduce disease severity and length of stay in intensive care.1 Critical illness is associated with challenges which stop an accurate assessment of nutritional needs. Patients may have chronic ill health, sarcopenia or obesity. In addition, sepsis may be associated with a stress response and catabolic state. The use of a published formula should guide prescription. A prescribed requirement of 25–30 kcal/kg/day is commonly used. A delivered volume which is close to that prescribed improves outcome.2
The aim of this prospective study was to compare the amount of nutrition delivered to that which was prescribed. We collected data on 15 patients who were considered to be established on enteral nutrition. Data were collected for three to seven days consecutively for each patient. We observed calories delivered, nitrogen received, and the type of feed. In addition, we collected data on the number of calories which were delivered using propofol. We collected data on reasons for, and the duration of, interruption. We had the constant presence of a nutrition specialist during weekdays on the critical care unit. We adhere to an enteral nutrition protocol which is reviewed daily. There was no feed – off period in our enteral feeding protocol.
The amount of calories delivered was 70–90% of that prescribed per patient. We observed that in a number of patients a large amount of the caloric intake was generated from propofol which contains triglycerides.
The average amount of protein delivered varied from 30 to 95%; however, we observed that it took a longer time to reach the prescribed protein requirement. It took four to seven days to reach the target in those patients who achieved more than 80% protein requirement.
The reasons for interrupted feed were delays around admission and commencing feed, concerns over abdominal diagnosis, delays while waiting for procedures or preparation for interventions. We acknowledged that very large gastric aspirates (>250 ml) were observed.
We conclude that despite enteral nutrition having a high profile in our department with protocols, educational programs and staffing in place, we do not achieve nutritional targets of enteral feed. We do not deliver enough calories or protein. We recognise that the triglycerides in propofol occasionally supply a significant proportion of calories. This suggests further underfeeding and nutritional imbalance. We acknowledge that hypertriglyceridemia is a risk.
References
Abstract 0093 – E-Poster EPM.175
Improving delivery of enteral nutrition in critical care patients: A quality improvement spiral introduction
S Clarke,
East Lancashire Hospitals NHS Trust
Nutrition is an important aspect of care of the critically ill patient. Prescribed daily enteral calorific targets are frequently not achieved, e.g. only 10% to 30% of the prescribed daily kCal of nutrition was delivered in the CALORIES study. Failing to deliver 25 kCal/kg/day may negatively influence patient outcome. We aimed to deliver greater than 85% of the prescribed kCal/day by initiating feeding at a higher rate rather than the previous stepwise increase, transitioning from an hourly based feeding protocol to a 24-h volume-based feeding protocol and increasing the permissible GRV from 250 ml to 300 ml (modified PEPuP protocol).
Abstract 0182 – E-Poster EPM.176
A service evaluation of enteral nutrition delivery in an adult intensive care unit background
Leeds Teaching Hospital Trust
Critically ill patients are at high risk of malnutrition. Delivery of nutrition to this population is complex, with large proportions requiring artificial nutrition. Nutritional requirements are often altered secondary to the metabolic effects of critical illness; however, even during periods of adequate nutrition, patients can experience loss of protein muscle mass due to catabolism. The importance of adequate nutrition has been well researched, early initiation of enteral nutrition (EN) (within 24–48 h of admission) has been associated with a reduction in infective complications and mortality rates. However, difficulties meeting requirements have also been highlighted. Causative factors include increased gastric residual volumes GRV's, enteral feeding tube migration or displacement, surgical intervention, fasting procedures, gastrointestinal (GI) disturbance, proning and some weaning processes. It is estimated that 60% of critically ill patients have some form of GI dysfunction leading to stoppages or delays in EN infusion.
- Determine the volume of EN received in comparison to the volume prescribed
- Establish reasons why EN was not received
- Evaluate the methods used to provide EN delivery
Reasons for missed feed: nasogastric tube (NGT) blockage (2.1%), blood in GRV’s (2.1%), NGT dislodgement (8.3%), fasting for procedures (8.3%), large GRV (10.4%), nil NGT (16.7%) and unknown (52.1%).
Abstract 0133 – E-Poster EPM.177
Does the NUTRIC score add to an established nutritional care plan?
C Ayres, D Howland,
Torbay Hospital
One of the considerations of the recent ASPEN guidelines1 is the use of the NUTRIC score to risk stratify patients and formulate a nutritional care plan. Currently, the intensive care unit (ICU) at Torbay Hospital does not use a nutritional screening tool. The purpose of the audit was to compare our standard nutritional care with the addition of NUTRIC scoring and to assess where changes could be made to improve nutritional provision.
Data were collected over a one-month period (March 2016). Patients aged 18 and above who had been admitted for at least four days were included, thereby allowing for data collection at 72 h of admission. NUTRIC and SOFA scores were calculated for each patient; patients’ observation charts were also evaluated to see if full feed rates/nutritional goals were achieved by the aforementioned time frame. Notes from patients not receiving full feed were reviewed to check if the appropriate interventions had been commenced (such as prokinetics). A number of other parameters were also evaluated within this time frame: consideration of feed; calculation of goal intake, target feed rate, target 24 h volumes, and total volume of administered feed in 24 h.
Twenty-six patients were included in this study, and their ages ranged from 35 to 90. It was not possible to obtain observation charts from two patients. Of the group, 46% (n = 12) were classified as high risk following NUTRIC scoring. Three (25%) high-risk patients were not receiving nutritional goal, with 14.3% (n = 2) in the low-risk group. The main reason justifying this was high gastric residual volumes (GRVs). Of the five patients not receiving goal intake, only three had a prokinetic prescribed.
To conclude, this audit has shown that risk stratifying patients does not add significant value to current nutritional care plans. However, it has highlighted areas of improvement such as: disregarding GRVs and prescription of prokinetics. Development of post-pyloric feeding services could also be investigated. Practical implementation of these improvements has been commenced locally and review of these changes will take place in six months.
Reference
Abstract 0037 – E-Poster EPM.178
Quality improvement project: Does regional citrate anti-coagulation improve dialysis filter life-span when compared to systemic heparin anti-coagulation?
Intensive Care Unit, University Hospital Crosshouse
In critically ill patients receiving continuous renal replacement therapy (CRRT), regional citrate anticoagulation (RCA) has been shown to improve dialysis filter life-span when compared to systemic heparin anti-coagulation.1–3 This project examined the dialysis filter life span in critically ill patients undergoing CRRT following a practice change from systemic heparin to regional citrate anticoagulation.
Run chart illustrating dialysis filter life-span in patients receiving systemic heparin compared with those who received RCA.
References
Abstract 0067 – E-Poster EPM.179
What are the barriers and enablers towards implementing the Sepsis Six at a large district general hospital? A qualitative study using the Theoretical Domains Framework
1Royal Cornwall Hospital
2City University
Each year 37,000 patients die from sepsis in the UK. Care standards in management of these patients are achieved in full in less than one in seven patients. One strategy to reduce this has been the ‘Sepsis 6’ care bundle, including basic but vital measures such as antibiotics and intravenous fluid resuscitation.1 Local uptake, however, has been very poor. This study aimed to identify barriers and enablers towards Sepsis 6 use at our hospital, in order to facilitate targeted behaviour change to improve performance.
The theoretical domains framework (TDF) is a tool designed to elicit and analyze beliefs affecting behaviour. Semi-structured interviews based around the TDF were conducted in a large district general hospital with a purposive sample of stakeholders including consultants, junior doctors and nurses from emergency department, medical and surgical admissions. Belief statements were identified through qualitative analysis. These were collated into a questionnaire where participants rated agreement with each as a barrier or enabler, and for importance, on a Likert scale. The questionnaire was then distributed to the entire target group from which the initial sample was taken. Median importance and agreement ratings were calculated for each role, clinical area, and overall. These were used to identify ‘relatively important, relative barriers’ and ‘relatively important, relative enablers’ as targets for behaviour change intervention.
A total of 1699 utterances were recorded and coded into 64 belief statements, which were collated into a 51-item questionnaire. One hundred and thirteen complete responses were obtained (44.3% response rate). Important barriers included beliefs around insufficient feedback and audit, poor teamwork and communication, concerns about risks vs. benefits of using the Sepsis 6 in certain patient groups, insufficient training, and widespread resource concerns, with insufficient staffing, time, and beds. Enablers included high confidence in knowledge and skills, widespread beliefs in the overall benefits of the Sepsis 6 and getting it done within an hour, widespread beliefs that identification and management of septic patients fell within everyone’s role, and that regular use of the Sepsis 6 made it easier to remember. Some beliefs were applicable for the entire group, whereas others were specific to roles or areas.
We successfully used a replicable method of eliciting barriers and enablers towards Sepsis 6 implementation at our hospital. This facilitates targeted performance improvement through behaviour change techniques. A multicentre study is underway to address external validity of barriers and enablers. This method could be used to address similar implementation problems within other hospitals.
References
Abstract 003 – E-Poster EPM.180
Adherence to local empiric antibiotic guidelines for respiratory infections by treating physicians in intensive care and effectiveness of this guideline
W Boer1,
1Hospital Oost-Limburg
2University Hospital Leuven
Reference
Abstract 0255 – E-Poster EPM.181
Close encounters of a third kind: Repeated supra-therapeutic ingestion of paracetamol (RSTIP)
Z Karim
Manchester Royal Infirmary
RSTIP is a similarly dangerous but poorly described third category of Paracetamol overdose, often due to therapeutic or iatrogenic misadventure.1,2 It is commonly – and potentially erroneously – managed as an intentional staggered overdose with four indiscriminate infusions of N-Acetylcysteine (NAC). Acute medicine remains pivotal in continuing care for Paracetamol overdoses; hence, a proper appreciation of RSTIP management will help select unsafe patients that are at higher overall risk than the single acute overdose.1 Using a wide literature search, this poster highlights and compares relevant evidence to re-educate acute physicians on this uncommon, little appreciated, yet important case presentation.
Five-part search strategy for evidence focussing exclusively on RSTIP: PubMed literature search; Toxbase guidelines review; Cochrane database review; UpToDate and Dynamed review; and a focussed Internet search. Evidence assessed and compared.
Search yielded nine papers, two guideline sub-sections and three educational reviews relevant to the study topic.3 There was universal agreement that clinical hepato-toxicity necessitates immediate NAC treatment. The dose threshold for RSTIP needing further assessment has no consensus evidence with a range of >200 to >75 mg/kg/24 h and >150 to >75 mg/kg/48 h. There is near universal agreement that in patients with whichever defined excess. Paracetamol levels >20 mg/L (>120 µmol/L) or ALT >50 IU/L necessitates NAC treatment; studies have not shown any adverse outcomes without these abnormalities.
Appropriate RSTIP management by acute physicians is important to balance the risks of NAC overtreatment with the dangerous outcomes of hepato-toxicity. Assessment thresholds are challenging to define; however when exceeded, case management appears fairly universal. A sensible take-home plan would be: In patients taking multiple doses over 24 h who have exceeded your local significant Paracetamol overdose threshold (or have hepatic risk factors) admit and treat those with Paracetamol levels >20 mg/L (>120 µmol/L) or an ALT >50 IU/L.
References
Abstract 0055 – E-Poster EPM.182
Continuing the audit spiral of the key service outcome measures according to the D16 service specification for adult critical care
P Frank1, S Vaughn2 and E Jackson2
1Royal Preston Hospital
2Blackpool Victoria Hospital
NHS England guidelines for the provision of intensive care services (ICS) state:
• There must be documentation in the patient record of the time and decision to admit to intensive care
• Admission must occur within 4 h of making the decision to admit
• Patients must be reviewed by a critical care consultant within 12 h of admission
Following an initial audit and completed action plan, we re-audited to assess whether services had improved and if further changes were needed.
We carried out a prospective audit of 50 critical care admissions to a 16-bedded district general critical care unit (CCU), including both level 2 and 3 patients. Each patient had a pro-forma completed. We analysed the data and compared our results to the previous audit.
We had two peaks in admission time, 15:00 and 19:00. Four were elective admissions from theatre, and consequently there was no documented time of decision to admit. Of the remaining 46 cases, a time of decision to admit was present in 29 cases. Only 1 of the 29 were admitted after 4 h of the decision to admit.
Of the 50 patients, 35 were reviewed by a critical care consultant within 4 h. Only eight patients were reviewed greater than 12 h after admission by a critical care consultant. The most common day for a delayed review by a critical care consultant was Saturday. Ten patients were reviewed by their parent speciality within 12 h of admission. In 18 patients, it took longer than 12 h and 22 patients were either never reviewed or the timing was not documented.
The CCU has improved performance in the key service outcome measure of critical care consultant review in under 12 h from 71% to 78% over a period of 12 months. Documentation of the time of decision to admit fell from 92% to 66%, despite a new time and date box being added to the patient admission chart. Recording a decision time is likely to pose difficultly for many units, and this approach has failed at our unit. An easier and more useful surrogate could be ‘Admission to critical care within 4 h of a request for a bed,’ as this would provide firm start and end points to the admission countdown. When a bed request is now made at our unit, the time of request is documented alongside the patient demographics. A further audit is now planned.
Abstract 0206 – E-Poster EPM.183
Use it, clean it, label it
1Broomfield Hospital, Mid Essex Hospital Trust
2Mid Essex Hospitals NHS Trust
Clinical equipment that are shared between patients, like ultrasound machines, can be conduits for infectious agents.1 The risk is enhanced when the equipment is used for invasive procedures in emergency situations, when lack of adequate cleaning between uses can be common. To prevent risk of hospital acquired infection, transmitted through contaminated equipment, a governance framework is necessary to set standards for decontamination and to ensure compliance with these standards.
We describe the initiation and effectiveness of a standard for the decontamination of ultrasound machines in a critical care environment.
A monthly audit of equipment decontamination practices was instituted in our institution (Broomfield Hospital) after observations of recurring residual blood contamination of ultrasound equipment. The audit was led by the Infection Prevention Link Nurse (IPLN), with a senior member of the medical team. The audit identified poor decontamination practices, with lack of responsibility and documentation regarding equipment use and care.
To address the problem of inadequate decontamination, a programme was instituted that included four components: (1) Establishment of an effective, but simple to use decontamination protocol, (2) education of the users about the problems with contamination and the use of cleaning standards, (3) supervision of compliance with regular feedback about performance to the users of the ultrasound machine, and 4. Empowerment of nurses and auxiliary staff to challenge inadequate decontamination practices.
Our approach has resulted in improved decontamination practices of ultrasound machines in the critical care environment. In our experience, consensus-led development of local protocols with establishment of clear ownership has been the major factors for the successful implementation. Maintaining the success rates in the face of protocol fatigue, however, requires constant vigilance and supervision.
Reference
Abstract 0214 – E-Poster EPM.184
Sedation hold practise in critical care
J Bramall, T Dessain and
Lister Hospital
References
Abstract 0233 – E-Poster EPM.185
Android tablet tool for continuous audit and quality improvement in ICU
A Cookson1, S Diamond-Fox1,
1Newcastle Upon Tyne NHS Foundation Trust
2Intensive Care Unit, Freeman Hospital
Intermittent, clinical audits, as part of a traditional Plan-Do-Study-Act cycle, can lead to modest improvements in the quality of care and remain a common approach in ICUs.1 Larger and more sustained improvements are possible with multi-faceted quality improvement programmes including continuous data collection and feedback to staff.2 While some large-scale projects have successfully used handheld devices to reduce the burden of data collection3 we are unaware of any available applications for individual ICUs. We report the development and piloting of an electronic tool for continuous audit of evidence-based care practices on a general ICU.
We chose to use a 7-in. android tablet for its small form factor and microSD slot for data transfer. A custom audit tool was created using Android Studio™ using JAVA.™ For data security reasons, the tablet was locked-out of wireless internet connection, and all other programs were disabled. The only means for data transfer from the audit program was via a pin-protected transfer to the microSD card. These data were then analysed using Microsoft Excel.
For the initial pilot, we decided to audit adherence to protective ventilation guidelines, prevention strategies for ventilator-associated pneumonia, and screening for delirium. Data were entered by junior medical staff during their afternoon shifts. The form was designed to allow data entry to take less than one minute per patient.
Data were recorded as discrete patient assessments; we did not attempt to record data every day, instead undertaking an ‘audit round’ on afternoons when time allowed, accumulating the results by calendar month. The basic fields recorded were age, sex, length of ICU stay and date of data entry. In addition, for ventilated patients we recorded maximum inspiratory pressure, tidal volumes 6–8 ml/kg ideal body weight, and 30 degree head of bed tilt. For delirium screening, we recorded compliance with morning and afternoon CAM-ICU screening and the results. We did not collect any patient identifiable data.
After an initial trial period in early 2015 and minor changes to the application, data collection started routinely in December 2015. Run charts were generated for each quality indicator allowing for ongoing assessment of our unit’s performance.
In conclusion, we developed and piloted an android tablet program specifically for continuous data collection of evidence-based ICU practices. We found it convenient, time-efficient and secure and believe it is a valuable alternative to paper-based, intermittent clinical audits.
References
Abstract 0246 – E-Poster EPM.186
Audit of management of hypoxic respiratory failure at a tertiary intensive care unit – Are we consistent, are we correct?
John Radcliffe Hospital
We retrospectively reviewed 15 months of medical and surgical admissions to our Adult General Intensive Care Units for evidence of hypoxic respiratory failure, identifying 117 patients according to admitting pathology and oxygen requirements. ‘Difficult to oxygenate’ was defined as an FiO2 of greater than or equal to .90 for four or more consecutive hours. Applying this criteria, 48 patients were identified all of which had P/F ratios of less than 13.3 kPa, indicating severe hypoxic respiratory failure and meeting the Extracorporeal Life Support Organisation (ELSO) guidelines for consideration of extracorporeal membrane oxygenation (ECMO).1 The electronic notes of this cohort were then reviewed, looking for evidence of discussion with the regional ECMO centre and the use of rescue measures for the difficult to oxygenate patient.
Of the 48 patients, one was excluded as it was clear that he was transitioning to end of life care. Of the remaining patients, 14 were documented as having been discussed with an ECMO centre, and six were retrieved by the specialist team. Thirty-two patients received neuromuscular blockade with either rocuronium or atracurium, and 12 were ventilated in the prone position. There was no consistent relationship between the utilisation of these rescue strategies, in terms of either severity of hypoxia or escalation from one to another. Lung-protective ventilation was also reviewed, using the Brompton quick calculator to ascertain tidal volumes of 6 ml/kg ideal body weight. Only four patients were ventilated at the suggested volumes throughout their period of hypoxia, whilst 17 had no tidal volumes recorded within the suggested parameters.
In conclusion, severe respiratory failure is a common occurrence in our intensive care units, but its management is inconsistent. There are, of course, limitations to the method of data collection as electronic documentation may not fully reflect the clinical situation and snapshots of ventilator practice may not fully reflect management over time. Despite this, however, there appear to be areas for development in consistency of management and adherence to best practice.
Reference
Abstract 0201 – E-Poster EPM.187
Out-of-hospital cardiac arrest: Evaluation of patient outcomes and impact on ICU resources
D Braganza1, W Davies1,
1Papworth Hospital NHS Foundation Trust
2School of Clinical Medicine, University of Cambridge
Recent guidelines recommend that patients resuscitated from out-of-hospital cardiac arrest (OHCA) and whose ECGs show ST-segment elevation (STE) receive immediate angiography with a view to primary percutaneous coronary intervention (PPCI), while those without STE should also be considered for angiography and PPCI.1 We set out to evaluate the impact of expanding PPCI provision on patient outcomes and ICU resource use.
Data on all patients admitted to a tertiary hospital via the PPCI pathway between January 2010 and December 2015 were analysed. Additional data from patients who sustained OHCA were analysed with respect to ICU admission and outcomes.
A total of 301 patients resuscitated from OHCA were admitted via the PPCI pathway during the study period. The number and proportion of OHCA cases increased from 5.5% (n = 38) of total admissions in 2010 to 8.4% (n = 67) in 2015. Mean number of co-morbidities per patient increased from 0.97 to 1.63 from 2010 to 2015. The proportion of OHCA patients diagnosed with definite myocardial infarction decreased from 92% (2010) to 60% (2015), while the proportion diagnosed with acute coronary syndrome increased from 0% to 31%. More required ICU admission (13% in 2010 to 58% in 2015) and support for two or more organ systems (11% in 2010 to 48% in 2015). Rate of survival to hospital discharge decreased from 100% in 2010 to 81% in 2015, and the proportions who were discharged without neurological deficit (92% in 2010 to 63% in 2015) and discharged home (95% in 2010 to 58% in 2015) both decreased. Declines in rate of survival to hospital discharge and proportion discharged to home were more pronounced among those admitted to ICU as compared to those not admitted to ICU.
Our findings suggest that more patients resuscitated from OHCA are being admitted via the PPCI pathway, with an increasing trend in patients with multiple existing co-morbidities. Despite higher rates of ICU admission and higher levels of organ support, smaller proportions achieve favourable outcomes. Our data suggest that expanding PPCI provision to more patients resuscitated from OHCA places increasing demands on limited ICU resources.
Reference
Abstract 0070 – E-Poster EPM.188
Implementation of an emergency laparotomy pathway to improve the quality of care delivery for the high-risk general surgical patients undergoing emergency laparotomy
S Body,
King's College NHS Foundation Trust
Comparison of indicators for emergency laparotomy patients pre-pathway, with a pathway and without a pathway.
Using quality improvement (QI) methodology, we designed and implemented a laparotomy pathway over several PDSA cycles, to incorporate NELA’s best practice recommendations. Process measures, such as pathway compliance, were assessed over two-week increments, along with five specific outcome measures, in order to assess the impact of the pathway (Table 1). In practice, the pathway is utilised by several healthcare teams throughout the peri-operative period and provides an aide memoire for peri-operative management and acts as a tool for improved communication amongst all clinicians. We have presented interim results of an ongoing QI project.
Initial success with pathway compliance (62%) was not sustained; therefore, the care bundle was further simplified and adjusted to be more users friendly. Recent results now suggest a sustained compliance of approx. 50%. Further analysis of outcome measures was performed for 43 patients identified during this time (11 April 2016–5 June 2016). Fourteen patients notes were unavailable. Interestingly, both groups demonstrated improvements compared with the pre pathway implementation cohort.
There is an increased uptake of key measures defined by NELA to improve the quality of care delivery for patients.1 Of particular note, an increase in risk stratification pre-operatively has allowed for shared decision making with patients to take place, along with forward planning of critical care services. This should reduce the pressure on service provision and unnecessary admissions of emergency general surgical patients. The results point towards the challenges of implementation when integrating multidisciplinary teams. Ensuring a cultural shift in the management of these high-risk emergency surgical patients will require further evidence of benefit before wholesale buy in is achieved from clinicians.
References
Abstract 0157 – E-Poster EPM.189
Myocardial infarction in intensive care units: A systematic review of diagnosis and treatment
D Atkinson1,
1Central Manchester Foundation NHS Trust
2Royal London Hospital
In view these uncertainties, we conducted this review to address two questions:
When the clinician suspects an ICU patient is undergoing an MI, what diagnostic tests are most helpful?
When an ICU patient is diagnosed with MI, what treatment is most appropriate?
Categories of papers.
It is possible to increase the sensitivity of regular monitoring for detecting MI used in many ICUs (e.g. clinical signs, 2 lead ECG monitoring, 12 lead ECG on demand and cardiac enzymes as prompted by suspicion of ischaemia) by several methods: continuous 12 lead ECG surveillance; regular 12 lead ECG and regular cardiac troponin (cTn) measurement unprompted by clinical suspicion.
Echocardiographic abnormalities were more common the greater the increase in cTn, and in one study were associated with increased likelihood of angiographic abnormality. However, lesions requiring interventional treatment remained unusual.
Suggested strategy for monitoring and diagnosis of MI.
References
Abstract 0007 – E-Poster EPM.190
Clinical outcomes of community-acquired severe sepsis after implement a simple severe sepsis fast track
P Chayakul, S Chusri, N Kiamkan, P Phunpairoth,
Songkhla Hospital
Abstract 0143 – E-Poster EPM.191
The relationship of preoperative regional cerebral oxygen saturation and PRE-DELIRIC scores in the prediction of postoperative delirium in cardiac surgery patients
C De Deyne, W Eertmans, F Jans,
Hospital Oost-Limburg
The aim of the study was to assess the relationship between preoperative SctO2 and the PRE-DELIRIC score.
Preoperative SctO2 was lower in patients with postoperative delirium (66%±3.9 vs. 68%±1.9 in non-delirious patients; p = 0.179). Nevertheless, no significant correlation was found between preoperative SctO2 and the PRE-DELIRIC score (r = −0.137, p = 0.406).
References
Abstract 0159 – E-Poster EPM.192
The challenges of delirium screening and management in practice: Experience from a mixed medical and surgical intensive care unit
S Benington and
Central Manchester University Hospitals NHS Foundation Trust
References
Abstract 0096 – E-Poster EPM.193
An audit of the depth of sedation and relationship with the development of delirium in intensive care
1Addenbrooke's Hospital
2MRC Biostatistics Unit
Delirium is a common complication of admission to critical care.1 Sedative drugs, and in particular benzodiazepines, may contribute to the development of delirium.2 We audited the depth of sedation in our ICU patients and examined the relationship between depth of sedation, sedative drug use and delirium.
Data were collected retrospectively for all patients admitted over a five-week period to our teaching hospital ICU. Depth of sedation was recorded hourly using the Richmond Agitation-Sedation Scale score (RASS), along with twice daily delirium screening (using CAM-ICU). Our guidelines recommend RASS 0 to −2 unless deep sedation (RASS ≤−3) is indicated during the hours directly following major surgery, if there is evidence of hypoxic respiratory failure (FiO2 ≥0.5) or if muscle-relaxants used. We recorded the cumulative daily dose of all sedative, analgesic, anxiolytic, antipsychotic and muscle relaxant drugs used. Statistical association between doses of common sedative drugs, duration of deep sedation and delirium was assessed using multivariate logistic regression. We also analysed the temporal relationship between sedative drugs, depth of sedation and delirium in individual patients.
Sixty-one patients were included of whom 24 (39.3%) developed delirium during their stay. The total time spent at RASS ≤ −3 was 36.0%. Of this deep sedation, 47.1% was considered to be indicated according to our guidelines. The drugs most commonly used for sedation in our unit were remifentanil and propofol. We found some evidence of association between the number of days of deep sedation and the development of delirium (odds ratio (OR): 1.50, p = 0.06) (Figure 1). There is also a suggestion of an association between delirium and the maximum daily dose (mg/kg/day) of propofol (OR: 1.03, p = 0.13) but not remifentanil (OR: 0.16, p = 0.51). For each patient, the CAM-ICU score throughout their stay was plotted together with the doses of propofol and remifentanil in order to show the temporal relationship between sedation and delirium (Figure 2). This illustrates that there are several patients who received deep sedation, which may have been avoided, for several days, and who subsequently developed delirium.
Our results show an association between duration of deep sedation and the development of delirium. This has led us to review the use of sedation and strategies to minimise unnecessary periods of deep sedation. Higher doses of propofol may be associated with the development of delirium. This merits further investigation and may support the increased use of alternative drugs such as alpha-2 agonists.
References
Abstract 0154 – E-Poster EPM.194
Prevalence of ICU-Delirium in a 60-bedded adult critical care unit
G Almeida,
Guys and St Thomas NHS
The CAM-ICU is a validated delirium screening-tool.
The audit aimed to establish the prevalence of ICU delirium using the CAM-ICU assessment tool in combination with a medical chart review.2
The diagnosis of delirium was based on either:
1. CAM-ICU, performed in patients with RASS >−4 and/or
2. medical chart review (electronic physiotherapy, nursing, medical notes).
The prevalence of delirium was 17% across the 173 patients.
Across the 30 delirious patients, the diagnosis was based on:
1. clinical opinion alone in 13 patients (43%)
2. positive CAM-ICU alone in eight patients (27%)
3. combination of positive CAM-ICU and clinical opinion in nine patients (30%).
The CAM-ICU was not performed due to a RASS<−4 during 61 patient-days (5%). In patients with RASS>−4, a CAM-ICU score was established during 890 (67%) patient-days.
1. Lower severity scores (APAHE II) compared to previous studies,1 due to the inclusion of our high-dependency units.
2. Daily sedation holds and early mobilization, carried out routinely in the department
3. The addition of the electronic chart review in the diagnostic algorithm yielding higher specificity
References
Abstract 0102 – E-Poster EPM.195
The relationship between zinc, selenium and vitamin D levels on outcomes in ICU
S McLaren, P Patel and
Hammersmith Hospital, Imperial College Healthcare NHS Trust
Zinc, selenium and vitamin D1 may be associated with adverse outcomes in critically ill patients. Proposed roles include the repair of oxidative stress-mediated damage from various processes such as sepsis induced organ dysfunction, and ischaemia-reperfusion injuries.2 We sought to ascertain if patients with lower levels of these three elements had longer length of stay in ICU or increased ICU mortality.
For all patients admitted to a 16-bed tertiary UK ICU during a four-month period Zinc, selenium and vitamin D were tested on day 1 of admission. Data was then collected on length of stay and mortality for that ICU admission. Pearson correlation coefficient was calculated to assess correlation between levels of each of zinc, selenium and vitamin D against length of stay. Unpaired t-testing was used to test if there was a difference in admission levels of these elements in patients who were successfully discharged from the ICU and those that died during their ICU admission.
A total of 118 patients were admitted during the period assessed, all of whom were tested for vitamin D on admission. Due to poor availability of the required blood bottles, zinc was tested in 49 of these patients and selenium in 55. Fifty-seven per cent were male (n = 67), and the largest ethnic group was white British comprising 35% (n = 41) of the cohort.
A significant correlation was found between length of stay and admission zinc (Pearson correlation co-efficient = 0.299, p = 0.041). However, no significant correlation was found between length of stay and admission vitamin D (Pearson correlation co-efficient = 0.143, p = 0.128) or Selenium (Pearson correlation co-efficient = −0.098, p = 0.486).
No significant difference was found in the mean vitamin D (p = 0.891), selenium (p = 0. 0.723) or zinc (p = 0. 0.744) between those who were successfully discharged from ICU and those who died during admission.
A significant correlation was found between zinc level on day of admission and length of stay in ICU. The cause of this remains unclear as does the role of zinc replacement in trying to reduce length of ICU stay, and requires further study. No significant correlation was found between selenium and vitamin D levels on day of admission and length of stay in ICU, nor between ICU mortality and zinc, selenium or vitamin D. A previous American study showed an association between vitamin D and mortality. The sample population in that study was 93% white1 and possibly the greater ethnic diversity in our sample is why similar results were not borne out.
References
Abstract 0146 – E-Poster EPM.196
The trials and tribulations of rapidly implementing a new enteral feeding protocol
C Federico, J Naisbitt, E Parkin,
Salford Royal NHS Foundation Trust
Patients in the intensive care unit (ICU) are often in a catabolic state due to critical illness.
Evidence suggests that an adequate caloric (25 kcal/kg/day) and protein (>1.2 gkg/day) intake during ICU admission can reduce the morbidity and mortality associated with critical illness.1,2 However, there are multiple barriers to meeting daily caloric and protein goals within intensive care.
Baseline audit data collected using a three-day snapshot method from our unit, which has 18 level three beds and cares for a mixed neurosciences/general patient population, suggested that:
• Only 61% of patients reached more than 85% of their intended caloric goal
• Only 54% of patients received more than 85% of their intended protein goal
As a result of this data, we initiated a quality improvement project with the aim of ensuring at least 95% of our patients consistently received at least 85% of their daily protein and caloric goal. To do this, we established a multi-disciplinary nutrition team and utilising root–cause analysis the team identified reasons patients were not meeting their nutrition goals. Following this, we proposed a rapid improvement program, using a series of Plan-Do-Study-Act cycles, tailored to our unit but building on the successes of others.3,4 Actions included:
• Changing from standard enteral feeds to a high-protein feeds
• Changing from rate-based feeding to a volume-based feeding protocol
• Increasing the tolerated gastric aspirate volumes and making a clear protocol for management of gastric aspirate volumes and managing impaired gastric motility
To ensure a smooth transition to the new protocol, we designed an intensive educational program, which included:
• 3 × 20-min face-to-face education sessions for two weeks for all staff
• An online calculator to facilitate easy calculation of nutritional requirements
• Education material regarding nutrition and patient centred outcomes (nibbles) were placed around the unit and laminated cards beside each patient’s observation chart as a quick reference guide.
Our initial results demonstrated a significant improvement with over 85% of patients meeting 85% of their individual caloric and protein goals. However, the instituted process of rapid change has not been without challenges, including:
• Timely adjustment in feed prescription to match changes in Propofol requirements
• Ensuring staff consistently re-calculate the rate of catch-up feeding after breaks
• Maintaining momentum and focus within the wider team to establish a cultural change
Our data collection is ongoing to assess the impact of further interventions and forcing functions. The results of our efforts will be presented at the SOA.
References
Abstract 004 – E-Poster EPM.197
How successful is gastric feeding in mechanically ventilated traumatic brain injury patients?
Addenbrooke's Hospital
Delayed gastric emptying (DGE) is common in mechanically ventilated traumatic brain injury (TBI) patients. Raised intracranial pressure (ICP) can cause a change in gastrointestinal motility and furthermore treatment with sedation, paralysis, osmotherapy and induced hypothermia can slow transit.1 Vomiting, high gastric residual volumes (GRVs), distension and constipation result in interruptions to gastric feeding and insufficient energy and protein delivery. Optimal nutritional delivery is considered >80% of prescribed, where underfeeding has been associated with worsened clinical outcomes.2
When gastric feeding fails due to DGE, prokinetics should be initiated and postpyloric feeding should be considered if these fail after 24–48 h.3 Nasojejunal tubes (NJTs) are often under-utilised in practice due to difficulties with insertion (lack of timely endoscopic placement), resulting in the perhaps unwarranted initiation of parenteral nutrition (PN). The use of laxatives to aid prokinesis and treat constipation should be initiated from admission (unless contraindicated).4
A retrospective service evaluation on our neurocritical care unit (NCCU) was undertaken to determine the following: (a) nutritional adequacy (energy and protein delivered versus prescribed in first 10 days of admission); (b) current treatment of high GRVs (>400 ml) and bowels with prokinetic and laxative usage; (c) consideration for NJT/PN. Thirteen mechanically ventilated TBI patients requiring gastric feeding and medical ICP management were analysed between February and April 2016.
Over the first 10 days of admission patients received 66% and 64% of prescribed energy and protein, respectively. Highest GRVs were observed on day six and seven: 38% of patients had GRVs >400 ml on day six and 46% on day seven. A positive correlation was observed between prokinetic usage and high GRVs: 62% of patients required single prokinetic treatment and 31% required dual therapy over these days. On average, patients did not have a bowel movement until day seven, suggesting a correlation with high GRVs. Most patients did not have laxatives started until day three of admission. Ninety-two per cent of patients were prescribed laxatives by day six, most requiring combination laxatives. Implementation of NJ feeding or PN was poor. Of the seven patients with high GRVs >400 ml despite prokinetics, only two patients had PN started and none had NJTs considered.
Gastric feeding failed to meet nutritional targets. Prokinetics were prescribed appropriately; however, treatment of constipation with the early initiation of laxatives could be improved. Maximising delivery of nutrition via the enteral route prior to initiating PN, with the use of NJTs for DGE is poor. Currently practicalities surrounding the insertion of NJTs restrict their use. With the development of self-propelling postpyloric tubes and studies reporting successful bedside placement,5 the use of NJTs in TBI patients on the NCCU requires evaluation to improve nutritional adequacy.
References
Abstract 0170 – E-Poster EPM.198
Nasogastric feed and propofol use in the nutrition of critically unwell patients: An acute trust experience
1Department of Anaesthetics, Colchester Hospital University NHS Foundation Trust
2Nutrition and Dietetic Service, Barking, Havering, and Redbridge University Hospitals NHS Trust
3Department of Anaesthesia and Critical Care, Barking, Havering, and Redbridge University Hospitals NHS Trust
Nutrition for critically unwell patients remains a concern despite improved outcomes seen from early feeding. In most critical care units (CCUs), there is a significant delay in initiating enteral nutrition that worsens the patient’s catabolic state, leading to organ dysfunction. Additionally, the use of lipid-rich propofol as a sedative provides a high-fat caloric intake, potentially worsening suboptimal nutrition. We describe the nutritional support provided to our CCU patients to determine the daily recommended calories and protein provided in the acute period and evaluate the role of propofol as a nutritional source.
Clinical notes were examined for patients admitted to an acute trust’s general CCU over 14 days including only those who were intubated and ventilated with no contra-indications to nasogastric (NG) feeds. The volume and type of NG feed and the volume of propofol used over a 24-h period were recorded to a maximum of seven days of admission. Daily calculated recommended energy intake (cREI) and protein intake (cRPI) were calculated by the CCU dietician. This was determined using their pre-admission nutritional status and ideal body weight with a target of 25–35 kcal/kg/day and 0.625–1.875g of protein/kg/day to create a required hourly rate of NG feed.
There were 37 patients identified, 13 women and 24 men, accounting for 122-patient days. There were 33 medical and 4 surgical patients with a median age of 64.5 (range: 22–85.4). NG feeds provided an average of 50.9% of the daily cREI for all patients. The average daily provision of protein was only 47.3% of the daily cRPI. In both cases, this was <15% on day 1 (D1) of admission but improved to roughly 60% on D7. Propofol provided an average of 36.1% of the daily intake with 68.7% of the calories on D1 and reduced to 29.1% by D7. When considering NG feeds and propofol together, our patients met 70.4% of the daily cREI. This was an average of 31.9% on D1 of admission and improved to 83.0% by D7.
Nutrition continues to be suboptimal within our CCU, reflected in the poor total calories and protein provided to our patients. This is worse on D1 due to the investigations requiring patient transfer and the lower priority in initiating NG feed. Additionally, propofol continues to be a significant source of calories. We have explored different strategies and have determined that a 24-h volume target would be better suited than our current practice.
Abstract 0194 – E-Poster EPM.199
Critical care junior doctor’s profile in a low middle income country: A national cross-sectional survey
A Abyadeera1, D Baranage2, S De Alwis3, AP De Silva4, S De Silva5, PG Mahipala6, A Padeniya5, D Ranasinghe7, PC Sigera2, K Thilakasiri, A Dondorp8, R Haniffa9, S Jayasinghe10 and S Munasinghe9
1University of Colombo
2Network for Improving Critical Care Systems and Training
3Office of Education, Training and Research Unit, Ministry of Health
4Network for Improving Critical care Systems and Training
5Lady Ridgeway Hospital
6Office of Director General of Health Services, Ministry of Health
7Government Medical Officers Association
8Mahidol Oxford Research Unit
9National Intensive Care Surveillance, Ministry of Health
10Department of Clinical Medicine, University of Colombo
Retention of junior doctors in specialties such as critical care is difficult, especially in resource-limited settings. This study describes the profile of junior doctors in adult state intensive care units (ICU) in Sri Lanka, a low-middle income country (LMIC).
This was a national cross-sectional survey using an anonymous self-administered electronic questionnaire. Self-perceived competencies for eight common ICU skills were assessed using a likert scale ranging from 0 to 10. The same scale was used to query which interventions would help develop their ICU and how “outsiders” to could contribute to such efforts.
In total, 539 doctors in 88 ICUs (5 ICUs declined) were contacted, generating 198 unique responses. Just under half of the respondents (93, 47%) work exclusively in ICUs. Junior doctors (150, 75.8%) had no previous exposure to anaesthesia and (134, 67.7%) had no previous ICU experience. One hundred and sixteen (60.7%) ICU doctors wished to specialize in critical care. However, only a few (10, 5%) doctors were currently engaged in any specialist training. Short-course training needs are shown in Table 1.
ICU junior doctors views on short courses.
Helpful interventions and outsiders help to improve ICU.
The vast majority (173, 88.2%) of doctors felt that the care provided for patients in their ICUs was good, very good or excellent, while 71 doctors (36.2%) would be happy to recommend ICU where they work to a relative with the highest possible score of 10 (on a 0–10 likert scale). Helpful interventions, including “outsider” help, as perceived by the doctors are shown in Figure 1.
Measures to improve training opportunities for these doctors and strategies to improve their retention in ICUs need to be addressed. The authors acknowledge Dr Dineshan Ranasinghe and Dr Kaushila Thilakasiri for their contribution for the study.
References
Abstract 0060 – E-Poster EPM.201
The use of ultrasound as an adjunct to peripheral venous cannulation by junior doctors
R Breslin, K Collins and
Blackpool Teaching Hospitals NHS Foundation Trust
Failure to achieve peripheral venous access when cannulation proves difficult can lead to patient harm through delays in care. Ultrasound is of value in such circumstances and is often essential in obtaining access, but despite this junior doctors are not routinely trained in its use. There is evidence that very junior doctors (e.g. FY1) are both capable of1 and benefit from2,3 acquiring basic bedside ultrasound skills early in their careers, but the independent use of such skills in clinical practice has yet to be demonstrated in this group. We have trained 20 FY1 doctors in the use of ultrasound as an adjunct to peripheral venous cannulation and following a post-training assessment of competence given them access to an ultrasound machine for use when attempting difficult cannulations.
Following two training sessions of only one hour each, 85% of participants passed an independent assessment of competence in the procedure. Participants reported the training to be enjoyable, effective and relevant to their clinical practice. Overall, self-reported confidence and perceived competence to operate an ultrasound machine, use it as an adjunct to peripheral venous cannulation and to obtain peripheral venous access in any patient without the need to escalate care to a more senior doctor improved. Over the next three months, we plan to collect data on the use of this skill in clinical practice by our cohort, as well as measure skill retention through reassessment of competence at one and three months post-training.
We have shown that FY1 doctors can be successfully trained in ultrasound-assisted venous cannulation. Through this ongoing educational research project, we hope to establish whether FY1 doctors can retain this skill, use it safely, effectively and independently in their clinical practice and whether this use benefits patients by avoiding delays in care.
References
Abstract 0229 – E-Poster EPM.202
Proxima: Time and motion study of arterial blood gas (ABG) sampling
M Grocott1,
1University Hospital Southampton NHS Foundation Trust
2Sphere Medical Limited
ABG sampling is carried out in intensive care units (ICU) every day. This is usually performed by a nurse, removing a blood sample from the patient, leaving the bed space, and taking it to an ABG machine for processing. This requires another member of staff to observe the patient in the nurse’s absence to maintain safety. There is a risk of blood splash when taking the blood sample, transporting and processing it, before the remaining blood is disposed of in a sharps bin.
The Proxima ABG analyser is a point-of-care (POC) device that enables ABG sampling to be performed at the patient’s bedside.1 It is a closed system, with no net blood loss as blood withdrawn for sampling is returned to the patient.
A time and motion study were undertaken to compare workload associated with using the Proxima versus standard ABG sampling, including safety aspects.
Twenty patients undergoing elective cardiac surgery were recruited, 10 were connected to the Proxima device and 10 were connected to a standard arterial line system. ICU nurses were studied taking ABG samples from each patient. The nurses were under continuous observation by independent data collectors to ensure accurate capture of all ABG sampling episodes.2 Episodes were timed, using a stopwatch, and associated data were recorded manually.
The average (mean) time taken to obtain an ABG sample result using the Proxima device will be compared with using the standard ABG sampling system. The average and typical time away from the patient's bedside, and the average and typical volume in ml of blood removed for analysis using a standard ABG sampling system will also be presented.
The results will be discussed in relation to safety issues and workload associated with ABG analysis.
References
Abstract 0131 – E-Poster EPM.203
ECMO and the antisynthetase syndrome
L Dyson,
University Hospitals of Leicester NHS Trust
We present two cases of patients with severe acute respiratory failure (SARF) secondary to pneumonia, who were managed with ECMO. Neither patient demonstrated significant improvement in oxygenation, despite reductions in inflammatory markers, until the administration of high-dose methylprednisolone and Rituximab. Both had antibodies consistent with the antisynthetase syndrome.
Antisynthetase syndrome is an autoimmune condition with antibodies attacking aminoacyl-tRNA-synthetases. Symptoms include myositis, polyarthritis, muscle weakness (particularly respiratory muscles), interstitial lung disease and Raynaud’s phenomenon. The interstitial lung disease can be progressive, and can cause fibrosis and eventual pulmonary hypertension. Antibodies identified vary, with Anti-Jo-1 being the most common, and include Anti-PL-7, Anti-PL-12 and several others. Treatment includes immunosuppression and symptom management with physiotherapy.
Despite both these patients presenting with a common cause of SARF, standard ICU treatment for pneumonia failed to deliver an improvement. When this occurs, we screen for a wide range of alternative diagnoses. In these cases, positive antibody results led us to treat with immunosuppression, with a dramatic improvement noted shortly thereafter. It is always important to keep unusual diagnoses (Zebra) as a possibility when recovery from common conditions (Horses) is not as pronounced as expected.
Abstract 0238 – E-Poster EPM.204
General anaesthesia to treat ventricular tachycardia storm: A case report
Royal Bournemouth Hospital
A 71-year-old man presented to a local hospital with nausea, diaphoresis and pre-syncope secondary to ventricular tachycardia (VT). He had a history of ischaemic heart disease, and subsequent episodes of VT had led to the insertion of an implantable cardiac defibrillator (ICD) in 2013.
For the next 48 h, he had repeated episodes of VT, some of which stimulated ICD shocks, depending on the rate of VT. Amiodarone and metoprolol were tried in an unsuccessful attempt to revert to sinus rhythm. He was subsequently transferred to CCU in our hospital for tertiary cardiology intervention.
Despite further amiodarone loading, magnesium and a lidocaine infusion, the patient continued to be in VT. Worsening cardiogenic shock developed requiring a noradrenaline infusion. At this stage, the decision was made to induce anaesthesia. Induction was performed as a modified rapid sequence induction with high-dose fentanyl. The patient was still in VT post induction, but after a DC shock sinus rhythm was restored and maintained. Sedation was continued for a further 48 h to allow instigation of regular amiodarone and bisoprolol. After successful extubation, the patient was transferred to CCU for consideration of ablation therapy.
Electrical storm is defined as three or more episodes of VT or VF in 24 h. Even with an ICD in situ, there is considerable morbidity associated with the phenomenon, with one prospective study finding electrical storm was an independent predictor of death.1
ALS guidelines suggest the use of amiodarone or DC shock. However, these patients have often had repeated appropriate firing of the ICD with little or non-sustained effect. The ACC/AHA/ESC 2006 guidelines2 suggest beta-blockade followed by amiodarone or procainamide. Revascularization should be prioritized if there is evidence of new ischaemia. Level IIb evidence is given for consideration of general anaesthesia. This is recognized as aiding suppression of sympathetic drive as well as to ameliorate the pain and distress from repeated ICD shocks. There have been only a few case reports3 where induction of general anaesthesia with propofol has allowed termination of the electrical storm.
Electrical storm is a rare process that can be difficult to control. Standard ALS protocols do not always work, and so a different approach is required. As well as focusing upon the cause, treatment must utilize sympathetic blockade, which may include general anaesthesia.
References
Abstract 0027 – E-Poster EPM.205&&
CVC position, arterial or venous? Time to be 100% sure
S George,
George Eliot Hospital
A 56-year-old man was admitted to the Intensive therapy unit (ITU) with acute kidney injury and rhabdomyolysis for hemofiltration and vasopressor support. He required noradrenaline via a central venous catheter (CVC), placed under ultrasound guidance in the right internal jugular vein (IJV) by a very experienced anaesthetist. After three days of treatment, it was discovered that the line was in an artery despite chest X-ray reports confirming a venous placement. A CT neck and thorax confirmed the line had gone through the right IJV straight into the right carotid artery at the site of insertion in the neck, with the tip ending in the aortic arch. The CVC was safely removed without the patient suffering any adverse consequences.
Documented failure rates of arterial puncture from CVC insertion range from 0.1% to 1.0% and result in subsequent placement of a large-bore catheter into an artery. Inadvertent arterial placement of a CVC may result in haemorrhage, pseudo aneurysm, stroke, or death. Pressure transduction is considered a highly reliable method to prevent inadvertent arterial cannulation. A manometer can detect arterial punctures not identified by blood flow and colour. Ultrasound guidance and pressure measurement are best seen as complementary rather than as alternative methods as they also are not definitive. Measurement of blood gases may be useful to detect inadvertent arterial placement in most cases but may be misleading in severe shock and arteriovenous malformation.
Arterial placement of a CVC can be disastrous and must be ruled out. Arterial placement of CVC may not be ruled out by the use of ultrasound or X-ray chest and to eliminate inadvertent arterial placement we suggest:
1. Connect CVC to a pressure transducer to visualise a wave form.
2. A venous blood gas may be an appropriate alternative in certain cases.
Abstract 0071 – E-Poster EPM.206
Case discussion: Severe traumatic brain injury (TBI) in a 32-week pregnant female. Is delivery of the foetus a component of managing raised intracranial pressure?
N Bunker,
The Royal London Hospital, Adult Critical Care Unit
A 32-year-old female, sustained blunt head trauma at 34 weeks’ gestation. The initial Glasgow Coma Score was 4. Airway obstruction was evident with oxygen saturations of 75%; pre-hospital rapid sequence intubation was performed. CT head scanning showed an acute left subdural haematoma with a small intraventricular left lateral haemorrhage (no midline shift). Multiple facial fractures and a hyoid bone fracture were noted. Foetal ultrasound and heart rate monitoring indicated that there was no immediate indication for delivery. An intracranial pressure (ICP) monitor was inserted and she was admitted to critical care.
In the subsequent 48 h, her ICP’s were well controlled and remained <20 mmHg. On day three, there were significant periods of raised ICP peaking at 40 mmHg. These were controlled with increasing sedation and intermittent paralysis. A repeat CT head showed evolution of the TBI and diffuse axonal injury. There were discussions between the intensive care and the obstetric team about delivery of the foetus; it was felt that the presence of the foetus and its effect on intra-abdominal pressure (IAP) may be contributing to her raised ICP. She underwent caesarian section and delivered a live foetus.
Over the next 24 h, her ICP continued to spike, with a peak at 70 mmHg; her pupils became fixed and dilated and she was given hypertonic saline and thiopentone. A repeat CT head was unchanged from previous but a decision was made to proceed to a decompressive craniotomy on clinical grounds.
Studies show that increasing the IAP significantly raises the ICP in neurotrauma patients.1 This is due to elevated intrathoracic pressure, which impedes cerebral venous outflow, increasing ICP and reducing cerebral perfusion pressure. In this case, the ICP remained high, despite conventional management. Reducing the abdominal pressure by delivering the foetus was considered optimal management. Pregnancy can be described as a chronically compensated state of raised IAP, but the effects in TBI are not well documented. One study has demonstrated a fall in IAP when intra-vesicular measurements were taken pre and post delivery.2 The use of a decompressive laparotomy as a treatment for refractory ICP in one study has indicated some benefit.3 This case illustrates the need for communication between intensive care and obstetric teams, optimising maternal well-being and balancing the risk to the foetus of premature delivery.
References
Abstract 0035 – E-Poster EPM.207
Intracranial haemorrhage – The feared complication of HELLP syndrome – Case report
S Baglyas, F Kovari and
North Middlesex University Hospital NHS Trust
The syndrome is a progressive condition and serious complications are frequent. The overall mortality of the pregnancies complicated with HELLP syndrome is around 1%, and cerebral hemorrhage or stroke to be the primary cause of death in 26% and the most contributing factor in another 45% of the deaths.
References
Abstract 0061 – E-Poster EPM.208
Time of death: Documentation critical
Southport and Ormskirk NHS Trust
Abstract 0068 – E-Poster EPM.209
Emergency management of adults with seizures – A quality improvement project
S Hinze1 and
1Department of Neurology, Great Western Hospital
2Bristol School of Anaesthesia
Generalised convulsive status epilepticus is fatal if not treated promptly.1 Options for treatment are multiple, with success closely associated with timely intervention, and adherence to protocol.2 Delayed intervention leads to treatment resistance and increased likelihood of refractory status. Patients who receive adequate first-line treatment are 6.8 (95% CI: 2.6–16.9) times more likely to have seizure termination.3 They also have shorter critical care and total hospital stays.4
Clinicians need to understand the different stages of status epilepticus and use appropriate medicines in a timely fashion.1 Use of an in-house protocol is highly recommended to provide the highest quality of care.4
In response to a case of status epilepticus I had been involved with, I reviewed case notes of these patients admitted to ITU 2014–2015. Management was compared to the SIGN guideline, as no local guideline existed. I also surveyed staff on the acute medical unit to assess their familiarity and confidence, prescribing and administering anti-epileptic medicines.
This identified that half of cases had incorrect use of benzodiazepines, with one-fifth of these ITU admissions due to iatrogenic respiratory depression. Phenytoin was incorrectly loaded 90% of the time, with patients most often undertreated. There was also a lack of availability of key medicines on wards and lack of confidence in those prescribing and preparing them.
To improve our management of status epilepticus in the emergency department, acute medical unit and ensure timely transfer to critical care, according to best available evidence, I liaised with key stakeholders from, Neurology, ED, AMU, ITU, and pharmacy to empower staff with a new, easy-to-use guideline.
The resulting guideline comprised a two-sided quick-reference sheet, supported by a full-text explanatory document. Our focus was to present immediately required information in an easy to follow format, while tackling the pitfalls identified during the note review. The recommendations are based on trial evidence where available, and where evidence is lacking, international expert consensus.
Time-referenced numbered stages encourage progression through the guideline, and tables with pre-calculated doses and administration instructions are provided to improve efficiency and reduce error. We also arranged for pharmacy to increase availability of suggested medicines, and introduced the guideline to staff at teaching events.
Our trust now has new guideline for the emergency management of adults with seizures, which we hope will help the trust overcome the problems identified. The guideline has been well received and we are evaluating its impact.
References
Abstract 0089 – E-Poster EPM.210
Significant improvement in ICU chest drain management following a completed audit cycle
D Sedgewick1, P Duggleby2 and B Millette2
1Oxford University Hospitals NHS Foundation Trust
2John Radcliffe Hospital
Management of chest drains has been greatly improved in the last several years following a rapid response report by the NPSA in 20081 and a national audit programme2 led by the British Thoracic Society (BTS). We aimed to audit practice in our institution using the 2010 BTS guidelines for management of pleural disease.3
We conducted a prospective audit over four months using an electronic proforma to capture all patients with a chest drain on the intensive care unit. The gold standard was 100% compliance to each of the criteria extracted from the 2010 BTS guideline. We then developed the departmental policy and electronic health record and actively engaged nursing and medical staff in multidisciplinary education to improve compliance with the guidelines. We then conducted a reaudit to assess improvement.
In the initial audit cycle, 23 chest drains were inserted in 17 patients. Written consent was only taken for two patients (8.7%). Fourteen patients were unable to consent due to reduced level of consciousness but did not have consent form four completed. The remaining seven patients had capacity but were not consented in either written or verbal form. There was deficient documentation leading to no record of ultrasound use in eight patients (34.8%) and no record of sterile technique in 12 patients (52.2%). Alarmingly, 14 drains (60.9%) were not connected to an underwater seal system, instead being connected to a drainage bag without a one-way valve. One patient (4.35%) drained more than 1.5 L in the first hour, placing this patient at risk of re-expansion pulmonary oedema.
After intervention, the second audit cycle demonstrated considerable improvement (see Figure 1 below). Nineteen drains were inserted in 15 patients, of which six (31.6%) had consent documented. Fourteen (73.7%) had use of ultrasound documented and 19 (100%) had sterile technique documented. Seventeen (89.5%) were connected to an underwater seal. No patients drained more than 1.5 L in the first hour.
These results demonstrate a considerable improvement in local management of chest drains particularly with regard to documentation and the use of an underwater seal attachment. Consent remains an issue for us and has also been consistently highlighted for improvement in national audits2. We will continue to make efforts to address this and maintain good practice in future audit cycles.
Percentage compliance with BTS guidelines pre and post intervention.
References
Abstract 0098 – E-Poster EPM.211
Protective lung ventilation: Can we do better?
Northwick Park Hospital
Low tidal volume ventilation at no greater than 6 ml/kg predicted body weight (PBW) has consistently been shown to reduce ventilator-associated lung injury and improve important patient outcomes in both ARDS and non-ARDS patients. However, calculating PBW requires the knowledge of a patient’s height. Actual height measurements are difficult to obtain as visual estimates, relatives’ knowledge and supine measurements are inconsistent.1 This audit assessed adherence to lung ventilation strategies using predicted body weight calculated by using different anthropometric methods of height estimation.
Prospective observational audit in January 2016 of 20 mechanically ventilated patients in an 11-bedded general medical and surgical ICU. Data were collected from observation charts for first 24 h of admission (excluding patients ventilated for less than 24 h). Height was estimated using measurements of ulnar length and lower leg (Chumlea method) and PBW calculated using the formula recommended by ARDSNet.2 The number of hours in 24 where tidal volumes >6 ml/kg based on PBW calculated using ulnar and lower leg length along with actual body weight (ABW) (as a comparator) were recorded.
The study found that during the 24-h measuring period, the patients were ventilated at greater than 6 ml/kg PBW for a median of 23.5 h (IQR: 19.75–24) and 23.5 h (IQR: 22–24) based on ulnar and lower leg estimates, respectively, and as a comparator 10 h (IQR: 3–19 h) based on ABW.
There was a significant difference in height calculation (range: −13 to +2 cm, mean difference: −3.34 cm, p < 0.05), PBW calculation (range: −11 to +3 kg, mean difference = −2.95 kg; p < 0.05) and 6 ml/kg calculation (range: −69 to +17 ml, mean difference: −18 ml, p < 0.05) by the two methods, with lower estimates found using ulnar measurement.
Ventilatory tidal volumes were consistently in excess of 6 ml/kg PBW in this cohort, and we demonstrated significant albeit small differences between the two anthropometric calculations of height using ulnar and lower leg length to calculate recommended maximum 6 ml/kg tidal volume. Further studies are required to assess whether or not these differences translate to important clinical outcomes. Introduction of disposable tape measures with integral height estimation, new observation charts to include PBW and further education of nursing and medical staff regarding lung-protective ventilation are also recommended.
References
Abstract 0155 – E-Poster EPM.212
A comparative survey of glycaemic control practice in the UK, Germany and Benelux
D Egbosimba and
Sphere Medical Limited
Upper thresholds for the target control level varied widely with strong differences between countries (Figure 1). While 54% of units reported a target range upper limit of 10 mmol/L or greater, this represented 70% and 40% of German and Benelux ICUs respectively. However, a majority of the German units with stricter threshold set this at 7.2 mmol/L, while no UK units had an upper threshold below 8 mmol/L.
While 99% of units used a BGA for measuring glucose samples, only an estimated 77% of measurements used this method (Figure 2). Despite concerns about suitability for this patient group,3 an estimated 19% of samples in 69% of ICUs are capillary blood tested on a glucose meter.
Challenges to implementing glycaemic control protocols were cited as: nursing dependency and time taken (19%); human errors particularly around feed changes (18%); hypoglycaemic episodes (14%); more liberal protocols than desired due to the risk of hypoglycaemia (13%); difficulties keeping patients within the desired range (8%); and lack of protocols or protocols being ignored (8%).
Upper thresholds of the target blood glucose concentration range by country. Estimated breakdown of tests by sample and measurement technique.
References
Abstract 0243 – E-Poster EPM.213
Fluid challenges: Who does what?
S Hudson1, V Bhardwaj1 and J Radhakrishnan2
1Broomfield Hospital Mid Essex Hospital Trust
2Mid Essex Hospitals NHS Trust
Fluid challenges are an important part of resuscitation; yet, there are wide variations in practice even within critical care units in Europe.1 Greater deviations can be expected in non-critical care environments but has not been quantified.
We proposed to assess the current practice in the administration and assessment of fluid challenges among practitioners of different qualifications and experience in a UK hospital.
The study was conducted at Broomfield Hospital, a district general hospital with tertiary specialist services. Current practice regarding fluid prescription and fluid challenges was assessed by an online questionnaire sent to all doctors and members of the trigger response team from March to June 2016. The ALS definition of the fluid challenge was used as the standard.
Simple descriptive measures were used to summarise the responses. Fisher exact test was used to compare contingency tables.
A total of 63 practitioners responded to the survey. Thirty-three (52.4%) of the responders were anaesthetists and critical care physicians, and the remainder from other specialities.
There was a wide variation in the volume of the fluid to be infused (250–100 ml) and the duration of infusion (5–30 min). Anaesthetic and critical care practitioners were likely to infuse larger volumes at a faster rate than medical practitioners.
Crystalloids were the preferred fluids for challenges with a clear preference for Ringer’s lactate as compared to normal saline (83.9% and 53.2%, respectively). Blood products and albumin were considered for fluid challenges by 14.5% (n = 9) and 3.2% (n = 2), predominantly in theatre and critical care environments.
Thirty-five (56.4%) respondents would use multiple physiological and clinical endpoints to identify the response to the fluid challenge. Members of the medical and surgical team were more likely to use single endpoints.
Our audit shows differences in fluid challenge strategies across different specialities in the hospital. These observed differences in infusion rate but not the selection of endpoints, are probably appropriate responses to the different clinical contexts faced by different specialities.
Our study shows that the variations observed in the FENICE study is exaggerated in non-critical care environments. A standard approach to fluid challenges with multimodal endpoints for efficacy should be encouraged.
References
Abstract 0028 – E-Poster EPM.214
An international survey of practice: The assessment of fluid overload and its management by intensive care specialists
R Bellomo1, N Glassford1, S Jones2,
1Austin Hospital
2Royal Darwin Hospital
3Royal London Hospital
Fluid overload is a common finding in the critically ill and there is increasing evidence that links cumulative fluid overload during admission to intensive care with detrimental outcomes.1 Fluid overload remains poorly defined,2 its optimal management has not been elucidated and there is a perception that clinical practice varies.
We distributed a structured online questionnaire to Intensive Care Specialists in the United Kingdom (UK) and Australia and New Zealand (ANZ) in order to document methods of assessment and management of fluid overload in the critically ill.
A total of 219 intensive care specialists responded to the survey, and 131 (60%) were from the UK. One hundred and twenty-eight (58%) worked in a tertiary hospital and 217 (99%) cared for critically ill medical and surgical patients, with smaller numbers also looking after surgical sub-specialty patient groups.
Graph demonstrating methods used to assess fluid status of critically ill patients. Graph demonstrating when Intensive Care Specialists would chose to use RRT ahead of diuretic therapy in fluid overloaded patients.
ANZ and UK ICU specialists use a combination of clinical examination findings, bedside tools and radiological features to assess fluid status, diagnose fluid overload and then initiate fluid removal in the critically ill (Figure 1).
Specialists targeted a clinician-set fluid balance as their most common strategy for managing fluid overload, followed by examination findings and physiological variables. Specialists were most likely to use renal replacement therapy (RRT) ahead of diuretic therapy for fluid removal outside of its traditional indications if patients were oligo/anuric or a trial of diuretic therapy had not generated an adequate response (Figure 2). Increasing vasopressor requirements was the most common reason clinicians would review their fluid removal strategy in a critically unwell fluid overloaded patient. There were no statistically significant differences in reported clinical practice between specialists in the UK and ANZ.
Self-reported practice with regard to fluid status assessment among ICU Specialists in the UK and ANZ shows that fluid overload remains poorly defined with variability in both management and practice, including the use of RRT.
References
Abstract 0079 – E-Poster EPM.215
Finding consensus in thromboprophylaxis prescribing
A Dushianthan and
University Hospital Southampton
The provision of adequate thromboprophylaxis for critically ill patients is a key component of ICU care; however, thresholds for starting and withholding chemical thromboprophylaxis vary amongst clinicians.1 Low-molecular weight heparins (LMWH) are more effective at preventing pulmonary embolism than unfractionated heparins (UFH),2 but they are less frequently administered to intensive care patients.1 We aimed to establish whether consensus around heparin prescribing existed amongst consultants on the General Intensive Care Unit at University Hospital Southampton and whether the use of LMWH could be increased.
We assessed consultant consensus through a two-part electronic survey. Part one required respondents to select chemical thromboprophylaxis for a series of case vignettes and part two asked for INR and platelet count thresholds in a range of scenarios.
Simultaneously, we audited the notes of 50 consecutive ICU patients to assess prescribing practice and identify patients where LMWH could safely be administered.
Ten Consultants completed the survey (67% response rate). Responses demonstrated a high degree of consensus, with an INR threshold of <2.0 and a platelet threshold of >50 for administration of heparin supported by 70% and 80% of respondents, respectively. Consultant responses to the case vignettes suggested prescribing practice fell within these thresholds and showed a strong preference for use of UFH over LMWH (47 vs. 15 responses).
We collected data from 210 patient days. Twenty-nine patients had at least one additional risk factor for venous thromboembolism and 12 patients had an identified bleeding risk at ICU admission. Chemical thromboprophylaxis was appropriately administered on 168 patient days and appropriately withheld on 20 patient days, giving a total compliance of 90%. There was again a strong preference for UFH over LMWH (126 vs. 27 patient days). eGFR was >30 ml/kg/1.73 m2 on 83% of patient days, meaning LMWH could potentially have been safely administered on 104 patient days where UFH was administered.
Consultants in our unit demonstrate a high degree of consensus around thromboprophylaxis prescribing and actual prescribing reflects this. There is significant scope to increase the use of LMWHs.
References
Abstract 0036 – E-Poster EPM.216
Translaryngeal tracheostomy: An audit and summary of 15 years’ experience
A Walder
North Devon District Hospital
Translaryngeal Tracheostomy
Fantoni technique
Why?
Safer – ventilated and protected airway during insertion
Less traumatic – continuous tamponade of tracheostomy track means less bleeding
Easier – less need for tube changes
Better outcomes – smaller neater scar
Preparation – same as blue rhino
Consent
Clotting
Sedation
Paralysis
100% O2, equipment, personnel
Any specific risks (anatomy, O2, head injury, etc.)
Stage 1 – Airway exchange
Using bougie change current ET tube to 5.0 mm cuffed tube.
Stage 2 – Retrograde wire
Needle in trachea between first and second tracheal rings
Wire fed up and out through mouth
Stage 3 – Dilation
Gentle traction and counter-traction with incision of the skin just enough to allow dilator through
Stage 4 – Reposition tube downwards
Using trochar slide tube (minus dilator) out, down posterior wall of trachea and down in the trachea
Translaryngeal Tracheostomy audit
15 years’ experience
About 1000 performed
No surgical tracheostomies
Two episodes of bleeding that required local stitches but did not compromise airway
One late tracheal stenosis needing surgery
One retained 4 mm end of wire – user error
A few slipping out of trachea on turning in early years
Abstract 0151 – E-Poster EPM.217
Plug the leak: A service review of extravasations on critical care
Sheffield Teaching Hospital Trust
Extravasation is a recognized complication following intravenous administration of medication,1,2 which may result in increased patient morbidity, cost of therapy, and length of stay.2 A review of extravasation events was undertaken following an increase of reported extravasations within our critical care (CC) units in 2013.
We aimed to assess the incidence of extravasation within our General CC Units, the drugs associated with the extravasation, the site and severity of extravasation, and the quality of documentation and management of extravasation injuries.
Extravasations occurring from November 2008 to December 2015 were retrospectively reviewed by searching Metavision, an electronic clinical information system. Extravasation injuries were graded and allocated a clinical severity score based on the documentation in notes. (Grade 4 = most severe, Grade 1 = least severe).
Eighty-one extravasation events were identified, giving a reported rate of 0.6 extravasations per 100 patients. Thirty-one per cent of which occurred at night. Infusions were associated with more of the extravasations than boluses. Mean patient age was 54 years; 56.8% were male and mean BMI was 26.7.
The mean severity scores (SS) for intravenous access situated in the forearm was 1.0, hand 1.08, ante-cubital fossa 0.88, feet 1.8 and central veins 2.5.
The highest SSs were noted with: noradrenaline and potassium chloride (SS 4), calcium chloride (SS 2.5), bicarbonate, insulin/dextrose and Hartmanns solution (SS 2).
Extravasations events were poorly documented in 40%, with good documentation seen in only 15.9% of cases. Cannulas were removed following extravasation in 39% of cases.
Central venous extravasation injuries, whilst the least frequent, had the highest mean SS. This is unsurprising as certain drugs are only administered centrally because of their potential to cause tissue injury. Extravasation injuries related to foot cannulas had a higher mean SS, perhaps due to the low caliber and relative fragility of many foot veins.
We aim to reduce the risk of extravasation by enhanced staff education, introduction of guidance for prevention and management of extravasation injuries, increased frequency of cannula site assessment, avoidance of foot cannulas for high-risk vesicants, and enhanced data collection regarding extravasation events using a Metavision-based form.
References
Abstract 0237 – E-Poster EPM.218
Implementing a new bundle to standardise sedation practices, ventilator weaning and delirium management in a tertiary critical care unit: As easy as ABCDE
C Day, T De Beer, I Johnson,
Queens Medical Centre
The ABCDE toolkit is a treatment bundle designed to improve the outcome of critical illness. The bundle consists of five interdependent components using an evidence-based approach to standardise patient care, reduce over sedation and shorten the duration of mechanical ventilation. Bedside safety checks and guidelines support members of the multi-disciplinary team to autonomously instigate spontaneous
The ABCDE bundle was introduced to the critical care unit at the Queen’s Medical Centre, Nottingham in phases from December 2015 with PDSA cycle implementation, of AB, CD then E. A process-based audit of ABCDE interventions and prescribing with assessment of appropriateness in clinical context took place six months after the final component was introduced. Fifty ‘patient-days’ were sampled on alternate days over two, week-long periods six weeks apart (eight days in total). To be included, patients had to have been admitted for at least 24 h so all the bundle processes could reasonably have been implemented. Comparison was made with a previous audit performed before the bundle was first introduced and interim snapshot audits.
Appropriate spontaneous awakening trials increased by 23% with sedation scores (assessed by Richmond Agitation Sedation Score) suggestive of increased wakefulness in the post-bundle study group (median RASS score −4 in afternoons pre-bundle to −3 post) with interquartile range analysis suggesting greater proportion of patients less sedated compared with the 2014 cohort. There was no documented evidence of a corresponding increase of spontaneous breathing trials. The choice of sedative agent was as per guideline in 79% of patients. The documentation of delirium assessment increased from one to three times daily and occurred more regularly than before implementation (77 to 92%). All anti-delirium prescriptions were as per guideline (100%). Early mobilisation occurred in 58% of eligible patients.
The ABCDE bundle has empowered members of our multidisciplinary team to progress patient care with autonomy assisted by bedside safety checks and guidelines in a tertiary centre critical care unit. Increases in spontaneous awakening trials, a general reduction in RASS throughout the day and improved delirium screening and management have been positive findings.
References
Abstract 0020 – E-Poster EPM.219
Retrospective audit cycle of the use of dexmedetomidine on intensive care
S Cooke, C Cowell and
Glenfield Hospital
Dexmedetomidine displays specific and selective α2-adrenoceptor agonism with a unique mechanism of action.1 We conducted a retrospective audit cycle on the use of dexmedetomidine on intensive care. Initial audit covered October 2013–October 2014 with 25 patients audited and re-audit April 2014–March 2015 with 18 patients audited.
In our initial audit, consultant initiation (61.5%) and documentation of indications (36%) was poor. Indications included agitation, biting the ETT, ETT/intervention tolerance and inappropriateness on sedation holds. Twenty patients (80%) were on ≥2 other sedative agents prior to commencing dexmedetomidine. Fourteen (56%) of whom had all other sedative agents ceased within 24 h of commencement and 16 (64%) within the target 72 h period. Eight patients did not show tangible improvement on dexmedetomidine, despite this it continued in seven patients. Of note, haloperidol use for agitation was infrequent and only four patients had received clonidine prior to commencement of dexmedetomidine.
Actions undertaken were a sticker for the prescription chart including indication and named consultant for initiation. Clonidine to be included in sedation guidelines as a less specific α2-agonist prior to dexmedetomidine use and haloperidol to be used for agitation.
Re-audit demonstrated significant improvement in documentation of consultant initiation (72%) and indications recorded (83%). Twelve patients (66%) were on ≥2 other sedative agents prior to commencing dexmedetomidine, with fewer on ≥3 agents. Fourteen (78%) patients had all other agents ceased within 24 h and 16 (89%) within 72 h. Haloperidol was used in 38% of patients receiving dexmedetomidine for agitation, but clonidine use as a first-line agent was low.
Experience with dexmedetomidine is growing and improvements have occurred. Adherence to consultant initiation appears to have ensured appropriate commencement and cessation of other agents. Clonidine use appears low in the re-audit, but this may in fact be due to an increased uptake and less patients progressing to require dexmedetomidine.
Abstract 0114 – E-Poster EPM.220
Introducing a new sedation policy in a large DGH: Before and after cohort analysis
Aneurin Bevan University Health Board
Reference
Abstract 0193 – E-Poster EPM.221
Managing severe delirium and agitation with dexmedetomidine in a level 1 trauma centre
G Marshall, R McKenna, S Ramsay,
Barts Health NHS Trust
A RASS (Richmond Agitation and Sedation Scale) within the range of −2 to +1 was considered favourable. If despite maximum tolerated dose of dexmedetomidine it was not possible to wean other sedation, we considered this a failure. Primary outcome was number of days to achieve desired RASS. Secondary outcomes were changes in dosage of sedative drugs, length of sedation, bradycardia and hypotension.
All patients showed reducing propofol and midazolam requirements over the duration of dexmedetomidine infusion (50% reduction within 24 h). Drug was stopped in one patient due to hypotension.
References
Abstract 0044 – E-Poster EPM.222
Adherence to surgical antibiotic prophylaxis guideline, still remains challenges; a brief report
1Iran University of Medical Sciences
2Isfahan University of Medical Sciences
Abstract 0087 – E-Poster EPM.223
Insertion rates and complications of central lines in the UK population: A pilot study
AVK Wong1, N Arora2, S Olusanya3, B Sharif4, RM Lundin4, A Dhadda4, S Clarke3, R Siviter3, M Argent4, T Szakmany4 and The ICNAP-1 Grou
1John Radcliffe Hospital
2Heart of England NHS Foundation Trust)
3Oxford University Hospitals
4Aneurin Bevan University Health Board
Abstract 0254 – E-Poster EPM.224
Intensively caring about glucose: The importance of managing hypoglycaemia
S Abbas and
Manchester Royal Infirmary
Interruptions to normal feeding, difficulties in assessing effects and new insulin resistance in critical illness, make maintaining stable glucose control particularly challenging in the critically ill patient. Being fairly easy to treat with supplementary glucose or insulin, there may be a tendency for clinicians to under-recognise the importance of avoiding excessive variability in blood serum glucose values.
In particular, hypoglycaemia – serum blood glucose less than 4mmol/L – warrants attention. Evidence shows the critically ill with unexpected episodes of low serum blood glucose have longer stays1 and higher mortality2 hence early recognition, treatment and forward planning warrant attention in managing these patients.
At a busy 40-bed intensive care unit in the North West, a retrospective, fact finding, data collection exercise was undertaken in late 2015 to investigate the unit’s causes of hypoglycaemia and their management. Relatively uncommon, 12 episodes of hypoglycaemia in level three patients showed a wide variation in causes of hypoglycaemia, measurement severity and pre-existing clinical condition. No patient had an obvious clinical change (0/12) and events were largely not documented (3/12). Two patients followed local guidelines for hypoglycaemia management; however, eight received a safe substitute overall. Of five level two episodes, all patients were type 2 non-insulin-dependent diabetics highlighting the heterogeneity in hypoglycaemic patient across critical care as a whole. Both documentation and management followed a similar pattern to the level three patients and suggests hypoglycaemia sometimes is not met with the urgency that their effects on long-term outcomes deserve.
This poster will use this project to highlight the need for intensive care units to re-examine their approach to the assessment and management of glucose abnormalities and re-energise a culture of hypo-awareness. It will also engage the broader discussion on the application and potential for continuous blood glucose monitoring as an additional, important and evolving method of monitoring the critically ill patient.3
References
Abstract 0014 – E-Poster EPM.225
Educating the team: Multidisciplinary approach to Safe Critically Ill Patient Transfer (SCRIPT) for registered ICU nurses
Hull and East Yorkshire NHS Trust
Patient transfers occur for a variety of reasons, intrahospital to scan/theatre or interhospital for specialist care or capacity reasons.
The sheer presence of a patient in the ICU, classifies them in the highest morbidity patient with the greatest potential for physiological disturbance from minimal distraction. This alone makes critically ill patients vulnerable to harm from transfer.
Protective factors of the ICU are lost when moving a patient, whether it is 10 yards down the corridor or 100 miles down the motorway. The classic transfer team is an ICU doctor of varying experience, nurse of varying experience and a porter or ambulance personnel of varying experience. All of which need to be fluent in several areas of the emergency patient care construct to establish and maintain patient safety integrity.
A concept that has been highlighted and driven within the NHS in recent years is the functioning of a multidisciplinary team, giving non-medics greater autonomy and an elevated platform from which to voice their opinion and in doing so has theoretically allowed responsibility to be dissipated. However, accountability is difficult to attribute to this autonomy if adequate training has not been delivered. Established staff often have experience to handle emergency situations outside the secure environment; however, those at the start of their career have yet to gain this experience, and both groups do not possess the intricacies of transfer physiology and subsequent disasters which puts added pressure on medical staff. This increases expectations on other team members, namely the accompanying doctor, expecting them to be fluent in their own job and multitask in developing their leadership, and teaching roles, in such an unstable and unpredictable situation. This implores the question of the actual integrity of the MDT and fairness, and questions the safety of a transfer.
So why do we not apply the same standards to nurses as we do to doctors? Several reasons, firstly attitudes to both accountability and education of nurses have yet to catch up to the work place expectations of inclusiveness in an MDT and secondly resources.
Currently, there is no existing regional training programme for the safe transfer of the critically ill patient aimed at registered nurses; however, there is an expectation for staff to have formalised training. This gap in education was identified, and consequently SCRIPT was developed to meet these requirements. SCRIPT is hoping to challenge and develop training for intensive care nurses to bridge these gaps by delivering an immersive simulation-based course designed around the current critical care network national competencies and based around the best practice guidance produced by The Intensive Care Society (2011). Who better to deliver this training, than the MDT. Critical care teacher trainer, with experienced ICU consultants and registrars and paramedics were engaged in the process, to develop and deliver a realistic, practically useful course that has been universally met with positive, excellent feedback from participants. “Fantastic courses, and have taken away many points to improve my future practice when transferring patients.”
Abstract 0150 – E-Poster EPM.226
Improving the safety of patients discharged from the ICU
L Bates, R Biggs, R Chambers, E Day, K Edwards,
University Hospital of Southampton
• To measure our baseline practice and to determine the number and severity of errors that occur related to the discharge process
• To determine how the introduction of electronic notes affected the errors that are made
• To introduce a standardised handover procedure
Abstract 0200 – E-Poster EPM.227
An audit of routine pregnancy testing using serum b-HCG levels in the intensive care unit at The Royal London Hospital
C Calderwood, E Fitchett, M-L Kong and
Bartshealth NHS Trust
A national audit in 2007 found that 26.7% of female admissions to ITU were of women of childbearing age (16–50 years), and 2.1% and 9.3% of these patients were ‘pregnant’ or ‘recently pregnant’ respectively.1
This audit was initiated in response to an index case of a 39-year-old female patient with a high BMI, who was admitted to ITU with a low GCS, seizures and severe hypertension. Despite multiple medical reviews, pregnancy testing was only performed two days into her admission.
This audit was done to assess the number of female patients of childbearing age who undergo pregnancy testing on the adult critical care unit (ACCU) at The Royal London Hospital. NICE guidelines recommend all women should have a pregnancy test pre-operatively.2 There is little guidance on pregnancy screening for ITU patients.
After the first audit cycle, measures involving increasing awareness and admission blood requests data set generation were introduced to increase the uptake of b-HCG testing for ACCU admissions and re-audit.
Patients were excluded from analysis if they were already known to be pregnant prior to admission.
After interventions, a re-audit was performed. Forty-seven female patients between 15 and 50 years were admitted to the ACCU between 4 April and 10 June 2016. Fifity-three per cent of our patients had evidence of either urine (14.9%) or serum (42.6%) b-HCG, two (4.2%) had both (Table 2).
References
Abstract 0220 – E-Poster EPM.228
Intracranial pressure monitoring protocols in UK neurological intensive care units
K Cooper and
James Cook University Hospital
The UK saw 162,544 admissions for traumatic brain injury (TBI) in 2013–2014.1
Intracranial pressure (ICP) monitoring is available within UK Neurological Intensive Care Units (NICU) and is maybe useful in early recognition of deterioration in unconscious TBI patients.
The Brain Trauma Foundation guidelines recommend ICP monitoring in patients with severe TBI and abnormal CT scans, and in those with severe TBI who are under 40 years old, have motor posturing or a systolic blood pressure of <90 mmHg.2
We wanted to assess if UK NICUs were using ICP monitoring as a standard of care in TBI. Our secondary objective was to ascertain if individual units had set protocols in place governing the usage of ICP monitoring.
The information was gathered via telephone survey. We called 20 NICU’s in the UK and spoke to the Registrar on Consultant available at the time and asked the following set questions:
1. Do you use ICP monitoring in head injuries?
2. If so, in which patients?
3. Do you have a protocol within your department?
ICP monitoring was used in 100% of units contacted and all units used ICP monitoring in traumatic brain injury. Two (10%) units had a specific TBI protocol; 30% (n = 6) had protocols which included ICP monitoring usage and the remaining 60% (n = 12) did not have a protocol or were unsure.
ICP monitoring is widely implemented in the management of traumatic brain injury across the UK; however, the circumstances in which they used is less well defined. Although ICP monitoring is widely used, it is clear that very few units have strict protocols governing its use.
References
Abstract 0221 – E-Poster EPM.229
Prescribing and therapeutic monitoring audit of the high cost antifungal drugs in a DGH
1Princess Royal Hospital
2Royal Surrey County Hospital
The audit aims to identify the adherence to local guidelines for appropriate usage of high cost antifungals including therapeutic drug monitoring (TDM).
The data sample was identified using a medication trace (through the pharmacy dispensing system JAC®) of the four high cost antifungals on the Hospital formulary. Patients who received any formulation of AmBisome®, Caspofungin, Voriconazole and Posaconazole were identified over the preceding two years (April 2014–April 2016).
Healthcare notes of the identified patients were then accessed for the audit data alongside the hospital’s pathology reporting system WinPath®. Parameters such as drug prescribing, endorsements, microbiology discussion documentation, TDM, documented dose banding for IV Voriconazole and AmBisome® and Child Pugh (CP) score in case of Caspofungin were assessed.
Seventy-two patient notes with a total 130 prescriptions were reviewed. The average compliance to the guidelines for choice (indication) or on Microbiology advice was 91.2% and is just short of the target 95%.
Compliance to the guidelines was 100% for the prescribing of dose, route and frequency but only 70% compliance with the drug form especially for Voriconazole as oral formulation exist in liquid and tablet form with differing bioavailability.
Percentage compliance for endorsement for indication was low for posaconazole (50%) and the percentage compliance for endorsement duration was low for AmBisome® (50%).
TDM was only necessary with voriconazole; however, 86% of all prescriptions for voriconazole did not have any documented results.
All three patients with liver derangement, who were prescribed caspofungin had documented CP score on the drug chart (100% compliance).
Patients with invasive fungal infections (IFI) have longer durations of hospital stay due to need for long periods of expensive inpatient therapy. Patients are also likely to suffer serious complications of the infections and toxicity issues associated with treatments. Despite cost associated with the use of these agents, the quality of prescribing remains poor.1
Education of prescribers regarding dose banding and TDM for relevant antifungals along with need to review lack of guidelines for certain specialities is recommended. Pharmacist need to challenge prescriptions where dosing is not in line with patient weight or dose banding tables (where relevant). Compared to previous audit in 2010, drug chart endorsements have improved; however, clear documentation and close liaison with microbiology for IFI are essential to help doctors prescribe safely and reduce any side effects.
Abstract 0207 – E-Poster EPM.230
Burn patients requiring intensive care due to infection
S Asai
Nagoya Kyoritsu Hospital
For the patients suffering from small burn injuries, an intensive care is not usually needed for the primary care, as they admit in the general ward on admission day. However, in rare cases, these patients’ condition could get worse while having a treatment in the general ward, and they can be forced to move to the intensive care unit (ICU) until the vital sign gets stable. In this time, it was examined the TSS complicated by small burn injury patients in adults. I report these rare cases with some our considerations.
Especially complicated by infection of TSS, quickly treatment is desirable, regardless of the age, TBSA and PBI.
PBD: post burn day; PO: post operation.
Abstract 0267 – E-Poster EPM.232
The fluid challenge technique: Standardization of fluid administration and assessment of fluid responsiveness in critically ill patients
Hollmann D Aya
St George’s University Hospitals & St George’s University of London
The variety of techniques may have an impact on clinical practice. The FC is frequently used in the process of early goal-directed therapy (EGDT) of high-risk surgical patients and septic patients. In high-risk surgical patients, EGDT reduces post-surgical complications and mortality.6 However, most haemodynamic studies in critically ill patients are performed in post-cardiac surgical patients, where this process is seen as controversial. Therefore, I performed a systematic review and meta-analysis7 of clinical trials of EGDT in cardiac surgery. Although mortality is not affected by this intervention, EGDT reduce morbidity and hospital length of stay in patients after cardiac surgery.
I then hypothesised that different doses of fluids may have different impact on Pmsf, and possibly different impact in venous return, CO and proportion of R and NR. I conducted a quasi-randomised clinical trial where 80 post-cardiac surgical patients were allocated to four groups: 1, 2, 3 and 4 ml/kg of crystalloids as FC infused in 5 min. I measured haemodynamics before and after the FC, including Pmsf-arm. I found that the minimal dose of crystalloids required to increase the Pmsf-arm by 14% was 4 ml/kg of total body weight. Importantly, the proportion of R increased along with the dose of fluids used, confirming my initial hypothesis (Figure 1).
Proportions of responders and non-responders by dose of crystalloids used for an FC. Proportions are compared across different doses of crystalloids with a chi-square statistic.
Similarly, a randomised clinical trial was also conducted in 80 septic patients using 2 to 5 ml/kg. No dose increased Pmsf-arm by more than 14% in these patients, and the estimation based on these data point out that about 9 ml/kg would be necessary to achieve a ΔPmsf-arm of 26.02% (95% CI: 14.07 to 39.96).
Finally, I was interested to observe the pharmacodynamics of an FC using 4 ml/kg, and also investigate the effect on the microcirculation and its relationship with the Pmsf. Thus, a study in 24 post-cardiac surgeries was conducted observing the effect on haemodynamics including Pmsf-arm and on sublingual microcirculatory variables using an incident dark-field camera. Data were recorded at baseline, immediately after the FC, 5 and 10 min after the fluid infusion. I observed that baseline values of microcirculatory parameters are inversely related to the overall effect of the FC on microcirculation (Figure 2). These changes in the microcirculation cannot be explained by the haemodynamic classification between R and NR. The FC with 4 ml/kg is able to increase arterial pressure in R (p = 0.01) and those changes are still present 10 min after the fluid infusion for CO (p = 0.01) and SV (p = 0.01). And finally, no relationship was found between the Pmsf-arm at baseline and the changes on Pmsf, with the overall changes of microcirculation in this sample.
Relationship between proportion of perfused vessels (PPV) at baseline and area under curve of PPV vs. time after an FC.
References
Abstract 0268 – E-Poster EPM.233
Echocardiographic physiological parameters to assess hemodynamic and cardiac electromechanics in the critically ill
G Tavazzi1, BV Patel2, A Barrades2, A Vazir2, D Gibson2 and S Price2
1University of Pavia, Anaesthesia, Pain Medicine and intensive care medicine, Foundation Policlinico San Matteo
2Anaesthetics, Pain Medicine and Intensive Care, Imperial College London
However, the current echocardiographic indices to assess the systolic and diastolic function in ICU, e.g. ejection fraction (EF) and diastolic indices, have been adopted from the cardiology population without being specifically validated in ICU and have some known potential limitations.4,5
We hypothesized that the investigation of physiological and reproducible echocardiographic indices allows an earlier recognition of cardiovascular impairment and a further understanding of hemodynamic status and underlying pathophysiological mechanisms.
We retrospectively analysed the data for 131 consecutive patients admitted to a cardiothoracic ICU and required a clinically indicated echocardiography between January and August 2012. Beside the standard echocardiographic parameters, we measured LV total-isovolumic time (t-IVT) and longitudinal function. t-IVT, which represents the time when the LV neither ejects nor fills(Figure1(a) and (b)), was previously seen to be the most sensitive indicator of perfusion mismatch and related to limitation of cardiac output and VO2 at peak stress during dobutamine administration in heart failure patients.6–8 t-IVT is a sensitive marker of hemodynamic status in patients in cardiogenic shock with adrenergic agents and positive pressure ventilation.8 Mitral and tricuspid annular plane systolic excursion (MAPSE/TAPSE) by m-mode echocardiography are an easy and reproducible method for the assessment of longitudinal systolic function9,10 (Figure 1(c) and (d)).
Patients were divided into three groups basing on the reason of admission: group 1 (55 patients were admitted for severe respiratory failure), group 2 (40 patients after cardiac surgery, with temporary pacing) and group 3 (36 patients for primary cardio-circulatory failure).
Clinical profile and therapeutic regimens of the patients were extracted from the patient data managing system.
Patients’ features are shown in Table 1.
t-IVT showed a strong inverse correlation with stroke volume (SV) and cardiac output (CO) above all the population (respectively p <0.001 (95% conf. interval 0.94–0.99)); p <0.001 (95% conf. interval 0.87–0.94) in all groups separately (Figure 2(a) to (c)):
Group 1 t-IVT and SV p 0.0001 (0.85–0.95); CO p < 0.0001 (0.04–0.38)
Group 2 t-IVT and SV p 0.006 (0.8–0.96); CO p 0.007 (0.04–0.62)
Group 3 t-IVT and SV p 0.0001 (0.87–0.96); CO p <0.0001 (0.18–0.58)
(a and b) Correlation between t-IVT and stroke volume (SV-A) and cardiac output (CO-B). (c and d) Kaplan–Meyer curve of t-IVT overall the population and splitting in respiratory and cardiac patients. (d) Correlation found between presystolic A wave (sign of right ventricular restrictive physiology) and respiratory and ventilatory parameters.
t-IVT showed a strong inverse correlation with MAPSE in the univariate analysis, in the overall population (p <0.0001) and splitting the population (group 1 p <0.001; group 2 p <0.0001; group 3 p <0.0001).
In the multivariate analysis, the MAPSE showed to be good predictors of the t-IVT.
In the univariate logistic analysis, the echo parameters CO, MAPSE and t-IVT were all predictors of the 28-day outcome. In the multivariate model, only the tIVTd (HR: 5.1, IC 95%: 2.2–11.8) and the MAPSE at inferior wall showed to be independent predictors of mortality (HR: 0.17, IC 95%: 0.028–0.97).
When analysing the discriminatory power of the tIVT as predictor of outcome, the Kaplan-Meier survival curve were >0.7 in all the groups and >0.8 in cardiac patients (Figure 2(d)).
The ROC for outcome for respiratory patients was 0.80 and 0.77 for SV and CO, respectively.
The ROC above the cardiac patients was group 1 SV 0.81 and 0.84 and, respectively, group 2 SV 0.82 and CO 0.84 and group 3 SV 0.81 CO 0.83.
The ejection fraction, end diastolic volume and fractional shortening show no correlation with outcome and with SV and CO when analyzed in either the complete cohort or separate groups.
Evaluation of right ventricular performance in ARDS cohorts and patients undergoing whole lung lavage
There is a paucity of data examining the role of RV dysfunction in ARDS.11 We studied 98 patients admitted with ARDS (37 on VV-ECMO) without acute cor pulmonale. We looked at systolic longitudinal function (TAPSE) and at diastolic parameters of right ventricular restrictive physiology (pre-systolic A wave)12and right ventricular t-IVT (measured at level of tricuspid inflow and pulmonary outflow).
The systolic dysfunction (TAPSE <2 cm) was found in 66.6% of patients and restrictive physiology found in 58%. Systolic dysfunction was related to higher Doppler estimated systolic pulmonary pressure (p 0.012), RV t-IVT (p 0.002) and hypercarbia (p 0.04) and restrictive physiology with pressure ventilation (p 0.001), hypercarbia (p 0.001) and with Doppler-estimated pulmonary vascular resistance (p < 0.0001) (Figure 2(e)).
We performed the same analysis on nine patients undergoing whole lung lavage for pulmonary alveolar proteinosis. This human model enables the assessment of RV function during alveolar flooding. Throughout the procedure, the patient underwent selective mono-pulmonary ventilation while the non-dependent lung was flooded with warmed normal saline solution in 1–2 L aliquots (total volumes up to 20 L), chest physiotherapy was performed, with the proteinaceous effluent drained through postural positioning. The protective mechanical ventilation was adapted to maintain adequate gas analysis. Echocardiography was performed at each tidal fluid ventilation. All the results were confirmed showing a strict relation of RV restrictive physiology with PEEP (p = 0.0001) and Plateau pressure (p = 0.0001) and between PV A wave and RV t-IVT (p = 0.001).
Specifically, LV t-IVT is the first and only echocardiographic parameter described in with hemodynamic status and to 28-day outcome in ICU. It is also the most sensitive overall index during chronotropic, dromotropic and inotropic manipulation.13
Amongst the complex pathophysiology of ARDS, previous results showed a high incidence of RV dysfunction with uncertain outcome data. However, all the papers published describe the relation between lung impairment and RV systolic longitudinal function.
We believe that, beside the standard echocardiographic indices, the use of more specific physiological echo parameters (RV t-IVT) should be advocated and further investigated in acute settings, such as septic shock or flow manipulation during extracorporeal support, for earlier recognition and optimization of cardiac dysfunction and facilitating a better understanding of cardiac pathophysiology in critically ill patients.
References
Abstract 0138 – New Generation E-Poster NG.001
Can an early prediction tool appropriately stratify patients into risk groups for developing critical care delirium?
1Cambridge University Hospitals NHS Trust
2Addenbrooke's Hospital
3MRC Biostatistics Unit
Delirium is a common complication of critical illness.1 Strategies can be placed to reduce the incidence of delirium.2 These may be of most benefit if aimed at patients at the highest risk. E-pre-deliric3 is a tool for predicting individual risk of developing delirium during an ICU stay. It is calculated from characteristics available at the time of admission. In this study, we piloted the use of E-pre-deliric in a large teaching hospital ICU.
Data were collected for all consecutive admissions over a five-week period. E-pre-deliric was calculated on admission (age, admission type and urgency, mean arterial pressure, serum urea and presence of respiratory failure, dementia, alcohol abuse or steroid use). Delirium incidence, defined by CAM-ICU, was recorded by retrospective analysis of electronic flowsheets. Predictive performance of E-pre-deliric was assessed by kernel density estimation (bandwidth = 4.5) and the area under the ROC curve (AUROC), using the statistics software R 3.3.1.
The study included 61 patients. Appropriately 88% of planned twice daily CAM-ICU scores were recorded. There were significant differences between our patient characteristics compared to that of the E-pre-deliric validation cohort; fewer males (42% vs. 59%), more medical admissions (54% vs. 34%), higher emergency admissions (85% vs. 60%), longer length of stay (median 5 days vs. 2 days) and a higher incidence of delirium (39.3% vs. 24.5%).
In our population, E-pre-deliric score had good specificity but lacked sensitivity (Figure 1). The AUROC of the E-pre-deliric score was 0.64 compared to 0.76 in the original study (where 1 indicates perfect prediction). Further analysis of our sample (Figure 2), comparing age, E-pre-deliric score and diagnosis reveals a core group of delirium-positive patients, admitted to the unit after either emergency surgery or unexpected complications of elective surgery. They were admitted to the unit ventilated and received an above average number of ventilator days. Despite the prolonged exposure to deliriogenic effects of sedation and ventilation, they had relatively low E-pre-deliric scores due to the initial elective nature of their intubation. The higher incidence of delirium on our unit is likely associated with our hospital providing alternative high-dependency locations for post-operative care, which divert away some lower delirium risk, shorter stay surgical patients.
Future study will use these results as a baseline, whilst we attempt to integrate a prediction score into the electronic hospital system, providing a delirium risk score that can be tailored to predict delirium within our ICU population.
References
Abstract 0147 – New Generation E-Poster NG.002
Continuous recording of respiratory rate to detect deterioration on general wards: Pilot study to evaluate feasibility for use with National Early Warning Score
S Ahmed1,
1Ysbyty Gwynedd
2Bangor University
Monitoring patients’ vital signs outside an ICU relies on nursing staff conducting checks at set intervals. However, if the patient deteriorates between monitoring times, there may be a delay in detecting the change in their condition. Continuous monitoring of vital signs, in particular heart rate (HR) and respiratory rate (RR), may provide a mechanism to alert doctors or nurses of an imminent serious clinical events and subsequently avoid serious adverse events. Feasibility of continuous monitoring within the UK context of the National Early Warning Score (NEWS)1 has not been evaluated.
VITAL II PLUS (REC reference: 15/WA/0377) is a prospective observational study conducted on two general wards in a district general hospital in North Wales. We compared recordings of respiratory rate by manual count with continuous recording from a wireless sensor during a five-day period. Allocation of devices was at the discretion of clinical teams.
Twenty-three patients were included in the preliminary analysis, 12 from a respiratory and 11 from a gastro-enterological ward with general medical patients. Ten patients had chronic hypoxia due to severe COPD or pulmonary fibrosis, nine patients had sepsis and three patients died. The mean length of stay was 14 days.
In total, there were 564 manual RR measurements and 43,894 sensor RR measurements. Frequency distribution of manually (Figure 1) and wirelessly (Figure 2) collected respiratory rates was compared. The proportion of high RR values that would have led to allocation of 3 points on the National Early Warning Score was significantly higher in measurements from the sensor than in manual measurements (41% vs. 7%, p < 0.0001). Mean variation of RR during a 6-h period was 21 (SD: 6).
Manual measurements of respiratory rate. Wireless sensor measurements of respiratory rate.
ITAL II PLUS has found marked differences between the manual and wireless sensor RR measurements: sensor-derived measurements were higher and more varied. Clinical implementation would translate in more alerts being generated.
Lower manual recordings might in part be due to measurements in resting patients. In order to assure comparability, a method of distinguishing between deterioration and ‘exercise’ needs to be implemented to overcome potential alarm fatigue relating to clinically insignificant alerts. Existing algorithms might need significant adaption if using continuous recordings.
Reference
Abstract 0166 – New Generation E-Poster NG.003
The INtensive Care wEights ProjecT – INCEPT: An audit of weight descriptors and their impact on treatment across nine ICUs in the Thames Valley
S Clark, COLLABORATIVE, A OXCCARE, R Mattin and Siviter
Oxford University Hospitals
We undertook a prospective multicentre project, to assess the methods by which patient weights are obtained in ICUs throughout the Thames Valley Region, which weight descriptors are used for dosing of treatments and their impact on those therapies.
Our primary outcome measure was the difference between recorded weight and IBW. Secondary outcome measures included tidal volumes and the accuracy of propofol infusion rates. Paired t-tests compared variables.
In total, 116 patients had a weight recorded. Seventy-one per cent were estimated or unknown. Mean measured IBW was 60.1 kg (95% CI: 58.13–68.01), mean ulnar length-derived weight was 65.1 kg (95% CI: 63.21–67.03). Mean recorded weight was 77.6 kg (95% CI: 74.76–80.99). There were significant differences between measured IBW and all recorded weights (p < 0.005).
Mean tidal volumes (ml/kg) were 8.1, (IBW, measured height), 7.6 (ulnar-derived IBW), 8.1 (IBW recorded height) and 6.5 (recorded weight). There was a significant difference between tidal volume delivered as calculated by ideal bodyweight compared to recorded weight (p < 0.025).
Mean propofol infusion rates (mg/kg/h) were 2.34 (recorded bodyweight) compared with 4.15 (measured IBW; p < 0.0001), 3.55 (ulnar-derived IBW; p < 0.02) and 4.11 (recorded height IBW; p < 0.0005).
Reference
Abstract 0186 – New Generation E-Poster NG.004
Capnography in ICU: The team needs to know
R Kearney and
Dumfries & Galloway Royal Infirmary
The 4th National Audit Project (NAP4) ‘Major complications of airway management in the UK’1 stated: failure to use capnography in ventilated patients likely contributed to more than 70% of intensive care unit (ICU)-related deaths; and that increasing use of capnography on ICU is the single change with the greatest potential to prevent deaths. Pascall et al.2 analysed litigious claims received by the National Health Service Litigation Authority relating to ICU from 1995 to 2012 and noted that respiratory/airway was the third commonest claim category overall and the commonest category relating to patient death. One-to-one nursing care of ventilated patients notices acute changes in patient condition. Nursing staff also have a key role in intubation on ICU and can be members of a resuscitation team. Improving ICU nursing staff’s interpretation of capnography has the potential to greatly reduce morbidity and mortality.
Baseline knowledge of all nurses trained to manage a ventilated patient was assessed using a questionnaire which consisted of six common capnograph traces: normal; oesophageal intubation; bronchospasm; increasing EtCO2; decreasing EtCO2; increasing FiCO2 (Figure 1). The same answers were displayed for each trace and the respondent marked all answers they considered correct. Small group teaching lasting 20 min was then provided by a trainee anaesthetist and the nursing clinical educator. It focussed on a '6-step approach' to understanding and interpretation. The same questionnaire was then administered.
56 baseline responses were gathered. Of these, none correctly identified all correct answers and, significantly, 31/56 failed to recognise oesophageal intubation. The effects of changing cardiac output and minute ventilation were poorly understood.
To date, teaching has been provided to 49 nurses. There has been a global improvement in responses (Figure 2). 38 respondents identified all the correct answers, and all respondents identified oesophageal intubation, the normal waveform and bronchospasm.
Airway complications in ICU result in significant morbidity and mortality, and costly litigation. They are commonly associated with poor understanding or usage of capnography. Nursing staff have the potential to be effective first responders to acute changes in patient condition. The baseline questionnaire demonstrated that their ability to interpret capnography traces was limited. Targeted training has resulted in a global improvement in knowledge and, consequently, has enhanced patient safety.
References
Abstract 0188 – New Generation E-Poster NG.005
IMproving PAtient Consenting Transparency for commonly performed ICU procedures: A quality improvement project (IMPACT-QI)
A Amaral, T Chimunda,
Interdepartmental Division of Critical Care Medicine, University of Toronto
Interviews with consultants and SDMs were performed to help development of an intervention tool. We have designed, and are piloting, a standard consent form and pictorial guide for SDMs for commonly performed intensive care procedures.
Reference
Abstract 0054 – New Generation E-Poster NG.006
Chest drains: Is the triangle of safety really safe? An ultrasonographic study of enshrined practice to improve patient outcome and safety
G Barker1, A Kingwill1 and
1Oxford University Hospitals NHS Foundation Trust
2John Radcliffe Hospital
Intercostal chest drain (ICD) insertion is a relatively common procedure but can be associated with significant complications including bleeding and organ perforation. Indications for insertion include management of haemothorax, pneumothorax and pleural effusions. The "triangle of safety (TOS)" is an often quoted site as being appropriate for ICD insertion, including in trauma patients by ATLS. The space is delineated by lateral border of the pectoralis major, lateral border of the latissimus dorsi, line of the fifth intercostal space and the base of the axilla intercostal space. Although ultrasound (US) guidance has been recommended to aid insertion, its use is far from established practice and landmark techniques and the TOS is still widely practiced.
To establish the safety profile of the TOS through the use of US to delineate underlying anatomy including the position of intercostal vessels.
Fifty consecutive patients on a general ICU underwent bilateral US examination of their TOS. The position of liver, spleen and heart in the respective TOS was noted. Doppler ultrasound was used to identify intercostal vessels within the TOS.
Overall, the heart, liver and spleen were visible in 60% of patients within the TOS. This percentage increased in the intubated and ventilated patient population. Intercostal vessels were visible in the minority of patients.
The British Thoracic Society and the National Patient Safety Agency UK has recommended ultrasound before inserting a drain for fluid. Our study found that the TOS is a misnomer exposing the patient to the risk of underlying organ perforation especially in the patient who is intubated and ventilated. It is therefore not safe and the practice should be abandoned. The routine use of real-time US to guide ICD insertion should be recommended to improve procedural safety.
Abstract 0010 – New Generation E-Poster NG.007
Single centre experience in high frequency percussive ventilation (HFPV) in early onset respiratory failure after cardiac surgery
Dept. Intensive Care, Critical Services, Ziekenhuis Oost Limburg
Mean Apache II score at the time of HFPV initiation was 22 (95% CI: 19-–26). The mean Euroscore II was 33.9% (CI 17,6–50,14). In six (35%) patients, surgery was elective, in 11 (65%) it was (semi-) urgent. Mean PaO2/Fio2 (mPFR) before HFPV was 100 (CI: 74–126), and significantly higher after 2 h (178 (CI: 124–232) (p = 0.01)) and after 24 h (207 (CI 120–296) (p = 0,03)). Overall mortality on ICU was 59% in this group (10 of 17). Mean cardiopulmonary bypass time (CBPT) was 272 min (CI: 226–318), reflecting often complex cardiac surgery. There was no correlation between urgency of operation and mortality. In non-survivors, means for EURSOCORE (43.8 vs. 19.8; p = 0.053), APACHE II (25 vs. 18; p = 0.062), CPBT (289 vs. 250 min; p = 0.42) and, surprisingly, mPFR after 2 h (243 vs. 148; p = 0.068) were higher, though none were significant.
Abstract 0245 – New Generation E-Poster NG.008
Practical acute care skills training (ACST) for newly qualified doctors – A peer-based model in a low middle income country
M Deshani1, P Mahipala2, T Stephens1, T Weerasinghe1, P Weeratunga3, S De Alwis4, A Padeniya5, D Ranasinghe6, K Thilakasiri1, A Beane1, P De Silva7, C Sigera7, A Dondorp8,
1Network for Improving Critical Care Systems and Training
2Office of Director General of Health Services, Ministry of Health
3National Hospital of Sri Lanka
4Office of Education, Training and Research Unit, Ministry of Health
5Lady Ridgeway Hospital
6Government Medical Officers Association
7National Intensive Care Surveillance, Ministry of Health
8Mahidol Oxford research Unit
9Department of Clinical Medicine, University of Colombo
Although undergraduate medical training lays the foundation for the knowledge and analytical skills that are required by junior doctors, there is a gap in the transference of this knowledge into the skills needed to act quickly and confidently in emergency situations. Good Intern Programme (GIP) in Sri Lanka has been initiated to bridge the ‘theory to practice gap’ of graduate doctors prior to their internship. This study describes the delivery and impact of an Acute Care Skills Training (ACST) programme for pre-internship doctors through a peer based training model, as part of the GIP.
A needs assessment was performed by an anonymous online survey of newly graduated doctors. The focus of the ACST programme was the recognition and management of common medical and surgical ward-based emergencies. Course content was developed by a multidisciplinary and inter-professional group including newly graduated doctors. The faculties of trainers were selected from the group of graduate doctors via a series of Train the Trainer programmes.
In total 902 (81%) pre-interns who completed the needs assessment survey stated that they would like to participate in a two-day practically focused ACST programme. Forty-eight per cent of them reported lack of confidence in interpreting key investigations for management of emergency situations, including ECG and ABG results.
The two-day training programme was conducted for small groups of up to 20 doctors over four months in late 2015. It was evaluated by 20 pre and post-course multiple choice test of 20 questions, a five station OSCE, a self-perceived skills assessment questionnaire and an anonymous candidate feedback form.
We delivered 17 courses over four months, training 320 participants by a faculty consisting of eight peer trainers. Post-MCQ scores were significantly higher when compared with pre-MCQ (p < 0.05). The post course self-assessments for all skills were significantly higher (p < 0.05) than the pre course self-assessments. The overall feedbacks from participants indicate a great majority strongly agreed that the course has improved their knowledge skills and confidence.
This experience demonstrates that it is possible to design and effectively deliver acute care skills training for pre-internship doctors in low middle income countries using a peer-based training model with support from more experienced local and overseas faculty. Peer learning could assist established medical schools in delivering interactive skills training necessary with minimal additional resources during their undergraduate training. Furthermore, similar peer programmes may have applicability beyond interns and beyond Sri Lanka, for honing essential practical skills.
Pre and post course self- evaluation of skills. Feedback for skills stations and scenarios from all course participants.
Abstract 0249 – New Generation E-Poster NG.009
Intensive care unit journey: Humanising the patient experience
C Evans1, G Fusari2, J Goldstone1, E Matthews2,
1University College London Hospitals
2Royal College of Art
NHS England outcomes framework identified five key domains to support staff to enhance patient care.1 Domain 4: ‘ensuring people have a positive experience of care’, is a challenge for all staff working in the Intensive Care environment.
Following collaboration with the Helen Hamlyn Healthcare Laboratory at the Royal College of Art, we are developing a novel patient-centred intervention, aimed at enhancing patients intensive care experience.
These key dimensions were further validated by the multi-disciplinary team and clinical psychologist focus groups, as well as through observation of the intensive care environment.
Through pre-hospital engagement of elective surgical patients due to be admitted to Intensive care, our tool identifies sensory preferences ranging from sounds and smells to relaxing videos and photos. The app will generate a mood board of sensory stimuli, personalised to the patient, which can be used throughout the day according to a schedule; this will help reduce boredom, provide structure to their day, while helping to orientate and humanise their hospital admission.
Reference
Abstract 0015 – New Generation E-Poster NG.010
Early non-invasive ventilation versus conventional oxygen therapy in immunocompromised patients with respiratory failure: A meta-analysis
K Flores, UK Gopez, N Maghuyop and
Philippine General Hospital
In these patients, we performed a meta-analysis on the effect of early NIV versus conventional oxygen therapy in reducing intubation rates and other important clinical outcomes.
The primary outcome of interest was intubation and MV rate. The secondary outcomes were ICU and all-cause mortality, ICU length of stay and duration of mechanical ventilation.
Patients on NIV had 1.18 days less stay in the ICU vs. oxygen group (95%CI: −1.84, −0.5 days). There was no statistically significant decrease in all-cause mortality between the two groups, RR 0.84 (95% CI: 0.63, 1.13), but this effect is heterogenous. After another sensitivity analysis performed specifically for this outcome, the results were homogenous and showed a 25% significant reduction in all-cause mortality in patients given NIV vs. oxygen therapy, RR: 0.75 (95%CI: 0.58, 0.96). There is no difference in the duration of mechanical ventilation between groups.
Differences in intubation and mechanical ventilation rates between those given early NIV versus oxygen. All-cause mortality benefit in the group given early NIV versus oxygen.
Abstract 0016 – New Generation E-Poster NG.011
Blood eosinophils as predictor for patient outcomes in COPD exacerbations: A systematic review and meta-analysis
J Magallanes,
Philippine General Hospital
Patients with blood eosinophilia had significantly shorter hospital stay compared to non-eosinophilic patients (mean difference: 0.68 days (95% CI: 1.09, 0.27)). Eosinophilic patients had significantly less frequent readmissions (odds ratio/OR: 0.69 (95% CI: 0.55, 0.87)), but there was no statistically significant difference in the one-year mortality rate (OR: 0.88 (95% CI: 0.73, 1.06)). Analysis showed a trend toward lower in-patient mortality among eosinophilic patients, although this difference is not statistically significant (OR: 0.53 (95% CI: 0.27, 1.05)). Furthermore, COPD patients with eosinophilia had significantly less need for mechanical ventilation during an exacerbation (OR: 0.56 (95% CI: 0.35, 0.89)).
Mean difference in hospitalization rates between the two groups. Difference in readmission rates between the two groups.
Abstract 0030 – New Generation E-Poster NG.012
Changes in Health-Related Quality of Life (HRQoL) after discharge from an intensive care unit: A systematic review
1Kadoorie Centre for Critical Care and Trauma Research and Education, John Radcliffe Hospital
2Nuffield Department of Clinical Neurosciences, University of Oxford
Yellow cells indicate where all studies were found to be different from their population norms (p < 0.05) (all were worse), blue where some studies differed and green where none. The figures in the cells are the number of studies where a statistically significant difference was found and the total number of studies with data at that time point and population norms.
Abstract 0032 – New Generation E-Poster NG.013
Using the EuroQol-5D to assess health-related quality of life in critical care patients during follow-up
K Ahmad1 and R Markham2
1Lancaster Medical School
2Consultant in Anaesthetics and Intensive Care, Royal Lancaster Infirmary
Many studies in the critical care literature place a focus on intensive care units (ICU) and in-patient mortality as a measure of primary outcomes. Consequently, the impact of any long-term sequelae is frequently overlooked.
There is clear evidence that follow-up clinics evaluating health related quality of life (HRQoL) in ICU survivors has a beneficial role.1 The introduction of NICE clinical guideline 83 (CG83) has helped profile the importance and promote the use of a standardised follow-up service. However, there has been limited development across the UK in implementing this. Only 27% of UK organisations offer a routine ICU follow-up service in line with NICE CG83 and 6.8% offer a post-hospital discharge rehabilitation programme.2
EQ-VAS and overall HRQoL (EQ-5D-VAS) at three months follow-up
To assess and compare the HRQoL across patient groups attending an ICU follow-up clinic using the EuroQol-5D-3L (EQ-5D-3L) questionnaire.
A retrospective study was performed at the adult ICU of the Royal Lancaster Infirmary, University Hospitals of Morecambe Bay NHS Foundation Trust. Patients with an ICU length of stay of ≥4 days between July 2007 and May 2016 were invited to attend an ICU follow-up clinic at three months post-discharge.
HRQoL was evaluated using the EQ-5D-3L questionnaire. The five dimensions of mobility, self-care, usual activities, pain/discomfort and anxiety/depression were converted in to a single summary weighted index (EQ-5D-VAS) ranging from −0.073 to 1, where 1 represents the value of full health. EQ-VAS records a patient’s self-reported health from 0 to 100. A total of 246 patients were included; medical (54.9%), trauma (6.1%), urgent surgical (31.3%), elective surgical (7.7%). The results for HRQoL are summarised in Table 1.
Trauma patients demonstrated a better health status when compared to the total population, median was 0.17 (IQR: 0.10–0.21). This was similar for the AKI subgroup in both EQ-VAS and EQ-5D-VAS scores; median was 88 (70–90) and 0.18 (−0.02–0.24), respectively. Cardiac diagnoses held the poorest health status; EQ-VAS, median was 75 (50–90), and EQ-5D-VAS, −0.02 (−0.07–0.08).
The EQ-5D dimensions showed that severe pain was reported in 67.1% of the total population and increased with age, 100% aged ≥80 years. Severe anxiety or depression was reported in 74.4% and increased with age, 100% aged ≥80 years. This was highest in cardiac patients (100%), followed by sepsis (82.1%), respiratory (75.6%) and invasive monitoring (75%).
All patients demonstrated an impaired HRQoL at three months post-ICU discharge. Severe pain, anxiety and depression are prevalent, showing a need for rehabilitation services to focus on addressing these states.
References
Abstract 0130 – New Generation E-Poster NG.014
The decisions we make – End of life on the ICU
S Anwar, P Jones,
Department of Peri-Operative Medicine, Barts Heart Centre, St Bartholomew's Hospital, West Smithfield
Although it is impossible to accurately predict the course of any one patient’s critical illness, there exists a plethora of data and scoring systems that can be applied to aid prognostication. In addition, physicians have their own inherent beliefs and biases, based on experiential data that have been shown to be highly accurate.1 Despite this potential for accurate outcome prediction, many incapacitated patients, with sequentially worsening prognoses, undergo prolonged periods of organ support without recovery. The absence of pre-morbid data of the patient’s own wishes makes decisions on the escalation of treatment particularly difficult.
The aim of this study was to use a simulated scenario to explore differences in end-of-life decision-making between intensive care physicians and lay members of the public, as potential future patients.
Reference
Abstract 0195 – New Generation E-Poster NG.15
Invasive or non-invasive ventilation at intensive care admission: A propensity score matched cohort study
S Fleming1,
1Univerity of Leeds
2Great Ormond Street
Patient characteristics were significantly different between MV groups. ‘IV-first’ patients had a significantly higher admission severity of illness (median PIM-2 score: 5.1% vs. 2.8%), serum lactate (median 1.9 vs. 1.6), and were more likely to have an arterial PaO2 value recorded (42.1% vs. 16.8%) compared to ‘NIV-first’ patients. Multilevel logistic regression revealed that PICU site explained 6.5% of the variation in first-line MV group (95% CI: 2.0% to 19.0%). The care area admitted from, primary diagnostic group, systolic BP, base excess (BE), BE polarity (positive or negative), and PaO2/FiO2 ratio at PICU admission were all associated with choice of MV modality.
In propensity score-matched analyses, adjusting for admitting PICU, primary diagnostic group, ethnicity, whether a blood gas was measured, admitting care area, age-standardised systolic BP, admission year, age and sex, receiving ‘NIV-first’ was associated with a significant reduction in mortality by 3.1% (95% CI: 1.7 to 4.6%), LOV by 1.6 days (95% CI: 1.0 to 2.3) and LOS by 2.1 days (95% CI: 1.3 to 3.0), as well as an increase in VFD-28 by 3.7 days (95% CI: 3.1 to 4.3).