Abstract
Increased advance care planning was endorsed at the start of the Coronavirus disease 2019 (COVID-19) pandemic with the aim of optimizing end-of-life care. This retrospective observational cohort study explores the impact of advanced care planning on place of death. 21,962 records from patients who died during the first year of the pandemic and who had an Electronic Palliative Care Coordination System record were included. 11,913 (54%) had a documented place of death. Of these 5,339 died at home and 2,378 died in hospital. 9,971 (45%) had both a documented place of death and a preferred place of death. Of these, 7,668 (77%) died in their preferred location. Documented elements of advance care planning, such as resuscitation status and ceiling of treatment decisions, were associated with an increased likelihood of dying in the preferred location, as were the number of times the record was viewed. During the COVID-19 pandemic, advanced care planning and the use of digital care coordination systems presented an opportunity for patients and healthcare staff to personalize care and influence end-of-life experiences.
Keywords
Key Findings
77% of people with a Coordinate My Care record with both a recorded place of death and a recorded preferred place of death died in their preferred location.
Less than one-fifth of individuals did not record a preferred place of death. Of these, a higher proportion died in hospital compared to those who did record a preferred place of death.
Documented elements of advance care planning such as resuscitation status, and ceiling of treatment decisions were associated with an increased likelihood of dying in the preferred place.
A lower socio-economic status was associated with a reduced likelihood of dying in the preferred place.
Background
Advance care planning involves dynamic discussions during which patient-specific goals and preferences are identified to try and ensure that end-of-life care aligns accordingly.1,2 Accessible, up-to-date documentation of end-of-life preferences, through digital means such as Electronic Palliative Care Coordination Systems, is fundamental to this process. 3
Place of death, and dying in the preferred place of death, are widely used as quality indicators for end-of-life care.4,5 Most people who express a preference say that they would prefer to die in their usual place of residence (home/care home) and most patients would not choose to die in hospital.6–8 Therefore, dying in a hospital and dying in a location that is not in keeping with individual patient preference can be considered factors contributing to poor end-of-life patient experiences and outcomes. Some people, however, do express a preference to die in hospital and others might actively choose to not express a preference.6–10
The Coronavirus disease 2019 (COVID-19) pandemic resulted in marked increases in excess deaths and an unprecedented impact on global health service delivery.11,12 In England during this time, more people died at home, and this trend has continued. 11 During the pandemic, advance care planning was widely advocated by clinicians and policymakers to improve patient-centered care and support end-of-life care coordination.2,13–16 Prior to COVID-19, advance care planning has been shown to reduce unnecessary emergency hospital admissions and increase the likelihood of an individual dying where they want to die.10,17–19 Little evidence has been published to date, however, about the relationship between advance care planning and place of death during the COVID-19 pandemic. The aim of this study was to describe the impact of advance care planning on the place of death during the COVID-19 pandemic using routinely collected data from an Electronic Palliative Care Coordination System.
Methods
Study Design, Setting and Participants
This is a retrospective observational cohort study of Coordinate My Care records for people aged 18+, who had consented to their data being anonymously analyzed for research and who died during the COVID-19 pandemic, between March 3, 2020 to March 5, 2021.
Coordinate My Care is a digital communication platform developed to support advance care planning for patients with complex, life-limiting conditions. It is the largest Electronic Palliative Care Coordination System in England, commissioned in London 2010 to 2022. 20 Once published, plans can be viewed and updated by patients, primary, and acute care services, including urgent care (NHS 111, 999) providers. Records contain individuals’ clinical and demographic information, as well as their wishes and preferences for care, including listing preferences for the setting of their care and place of death. 21 Coordinate My Care is integrated with the National Health Service central IT infrastructure “Spine” which allows health and social care information to be shared securely and enables alignment of key data such as date of death. Information on a persons’ actual place of death is reliant on a healthcare professional manually updating individual patient record data.
Data Analysis
The study population was described using descriptive statistics in terms of patient demographic, socio-economic and clinical characteristics, documented elements of advance care planning (such as cardio-pulmonary resuscitation status, ceiling of treatment, and preferred place of death) and the total number of times that the records were viewed by urgent and nonurgent care services within the 30 days before death. Records without a recorded actual place of death and preferred place of death were also examined.
Forced multivariable logistic regression was used to develop models of factors associated with the co-primary endpoints:
− Death in the preferred location. Where preferred place of death matched actual place of death. Individuals who did not have both a recorded preferred place of death and a known place of death were excluded. − Death in hospital. All records with a recorded place of death were included.
All factors believed to be associated a priori were included in the final model.8,22 Geographical location (based on Clinical Commissioning Group, clinically lead local area organizations tasked with delivering NHS services during the study time period), timing within the first year of the pandemic (stratified by surges/waves in COVID-19 cases indicated by Public Health England Data
23
), and time between record creation and death (days) were also included in the models. The postcode from the patient's place of residence was used to calculate the index of multiple deprivation (IMD) for England as a marker of socio-economic status. IMD represents a combined measure of deprivation based on seven domains, each reflecting different aspects of deprivation experienced by individuals living in an area. These areas are classified into quintiles with IMD Quintile 1 representing areas with the highest deprivation and Quintile 5 areas with least deprivation.
24
Data were analyzed using STATA v16 (StataCorp). The 95% confidence intervals which do not overlap the null value (odds ratio [OR] = 1) were considered statistically significant. 25 Data were reported according to the RECORD guidelines. 26
Regulatory and Ethics Committee Approval
This study was approved by NHS Health Research Authority (Integrated Research Application System reference 294940). The research was exempt from NHS Research Ethics Committee review as it involved the secondary use of anonymized data collected in the course of normal care. 27
Results
In total, 21,962 people with a Coordinate My Care record died during the study period Table 1. Most individuals were aged over 80 (67.5%, n = 14,822), with a WHO performance status of 4, indicating they were completely confined to bed or chair, unable to carry out any self-care and reliant on others for their care needs (57.0%, n = 12,512).
Summary of Results.
Abbreviations: IMD: index of multiple deprivations; COVID-19, coronavirus disease 2019; ACP, advance care planning.
†Other diagnoses include neurological, gastro-intestine, renal, endocrine, haematological, mental health, musculoskeletal, vascular, COVID-19, “Other.” *Not included in the model due to collinearity. Variables with significant missing data: Ethnicity, capacity to make and communicate resuscitation decisions, receipt of care, family support,
The majority of individuals (85.5%, n = 18,799) had a “Do Not Attempt Cardio-Pulmonary Resuscitation” recommendation recorded which had been discussed with the patient (67.2%, n = 14,678) and their family (74.3%, n = 16,317).
Most individuals wanted to be cared for and die in their usual place of residence (home/care home, 86.0%, n = 18,890 and 71.7%, n = 15,75, respectively). In total, 2.1% (n = 454) wanted to die in hospital. In total, 19.5% (n = 4295) did not record a preference for place of death.
Overall, 54.4% (n = 11,913) had a documented place of death. Of these 5339 (44.8%) died at home and 2378 (20.0%) died in hospital. In total, 45.4% (n = 9971) of all the records had both a documented place of death and preferred place of death. Of these, 76.9% (n = 7668) died in their preferred place of death and 16.3% (n = 1621) died in hospital. In total, 31.8% (n = 757) of those who died in hospital did not record a preferred place of death.
Individuals who did not have an actual place of death documented on Coordinate My Care (45.8%, n = 10,049) were more likely to live in a care home (65.7%, n = 6600 vs 27.9%, n = 3319 of those with a recorded place of death), and less likely to have a cancer diagnosis (26.1%, n = 2,621vs 41.5%, n = 4948) (Appendix II). A smaller proportion of records for the individuals without a recorded place of death were viewed by urgent and nonurgent clinical services.
In the multivariate regression modeling, dying in the preferred location Figure 1A was associated with a “Do Not Attempt Cardio-Pulmonary Resuscitation” status (OR: 1.51, confidence interval [CI] 1.17-1.93), a Ceiling of Treatment for “Symptomatic Treatment” (OR: 3.52, CI 2.77-4.50), documented family discussions regarding resuscitation status (OR: 1.51, CI 1.33-1.72) and at least two nonurgent care record views in the 30 days before death (OR: 1.27, CI 1.13-1.43).
(A) Forest plot of odds ratios from forced Logistic Regression modeling on the determinants of dying in the preferred place of death. (B) Forest plot of odds ratios from forced logistic regression modeling on the determinants of dying in hospital. Full results including p values are presented in Appendix III.
The same variables were found to be associated with a decreased likelihood of dying in hospital Figure 1B : “Do Not Attempt Cardio-Pulmonary Resuscitation” status (OR: 0.6, CI 0.49-0.74), a Ceiling of Treatment for “Symptomatic Treatment” (OR 0.23, CI 0.18-0.29), documented family discussions regarding resuscitation status (OR 0.66, CI 0.59-0.75) and at least two nonurgent care record views in the 30 days before death (OR 0.75, CI 0.67-0.84).
Individuals from more deprived areas were less likely to achieve their preferred place of death (OR 0.65, CI 0.54-0.79) and more likely to die in the hospital (OR 1.34, CI 1.11-1.62).
Discussion
Population-level data have identified a change in where people died since the onset of COVID-19 with a sustained increase in home deaths.11,28 Our large study of routinely collected data from an Electronic Palliative Care Coordination System (Coordinate My Care) provides a unique insight into the end-of-life care preferences and experiences for people in a large urban area who died during the first year of the pandemic.
The data show that during this time most people with serious illnesses, who expressed a preference and had this recorded in the electronic record, wanted to die and be cared for at home/care home. Our analysis identifies that patient-level decision-making and preferences about place of care and death, cardio-pulmonary resuscitation, and ceilings of treatment were associated with where people died, supporting the concept that advance care planning contributes to improved end-of-life outcomes and experiences. 10
Factors previously identified as being associated with place of death, including performance status and primary diagnosis,8,22 remained significant during the pandemic. Inequalities in palliative care services for those of lower socio-economic positions are well described. 29 In this study we identified an association between socio-economic position and place of death. Patients who lived in areas of higher socio-economic deprivation, as represented by the IMD, were more likely to die in the hospital and less likely to die in the location of their choice. Additionally, we found that achieving the preferred place of death was associated with other modifiable factors, including the discussion of resuscitation recommendations with family. People whose electronic record was reviewed more frequently before death were more likely to die in their preferred place and less likely to die in hospital, highlighting the relevance of accessible care records.
Given the increased national emphasis on the provision of and engagement with advance care planning, 16 concerns regarding the quality of the process employed have been raised. 14 When considering the quality of end-of-life care, both patient-focused and service-related outcomes must be considered. Our study indicates that, even in these unprecedented circumstances, digital advance care planning communication platforms contribute to end-of-life care which is in keeping with patient preferences. This provides evidence to support the continued use and development of digital communication systems for patients and health services at the end of life.
Strengths and Limitations
This is the first study to examine the relationship between advance care planning and place of death during the COVID-19 pandemic. We were able to include patient-level information including family involvement in serious decision-making, where evidence of their role is still nascent. 30 Enabling and empowering patients to die where they choose may contribute to improved end-of-life experiences. However, we could not identify from these data whether patient experiences of the care received corresponded with expectations of the patients and their families. Place of death was missing from just under half of the initial cohort, limiting the number of records that we could include in the modeling and introducing the potential for selection bias. Digitally linking the patient-centered data in Coordinate My Care with other routinely collected national datasets would strengthen the reliability of these findings and enable the inclusion of other missing predictor variables such as ethnicity.8,22 Research from the USA indicates individuals from ethnically diverse and minoritized communities have been disadvantaged in terms of advance care planning.31-34
Conclusion
Place of death, and whether it is in keeping with the wishes of the patient, and his/her family, is often considered a quality indicator for end-of-life care. 35 Even in the uncertain setting of the pandemic, we found that advance care planning, including documenting and digitally sharing patient-level decision-making and preferences about place of care and death, cardio-pulmonary resuscitation, and ceilings of treatment were associated with where people died and present an opportunity to modify and enhance end-of-life care experiences and outcomes.
Footnotes
Appendix I
The RECORD statement—checklist of items, extended from the STROBE statement, that should be reported in observational studies using routinely collected health data.
Appendix II: Summary of Patients With and Without a Recorded Place of Death
| Outcome | All deaths n = 21 962 n, (%) | All deaths with place of death recorded n = 11 913 (54.2%) n, (%) | All deaths without a place of death recorded n = 10 049 (45.8%) n, (%) | Proportional test P value |
|---|---|---|---|---|
| Patient Characteristics | ||||
|
|
||||
| Female | 12,090 (55.1) | 6487 (54.5) | 5603 (55.8) | .052 |
| Male | 9872 (45.0) | 5426 (45.6) | 4446 (44.2) | .052 |
|
|
||||
| 18-39 | 107 (0.5) | 70 (0.6) | 37 (0.4) | .020 |
| 40-49 | 308 (1.4) | 207 (1.7) | 101 (1.0) | <.001 |
| 50-59 | 925 (4.2) | 606 (5.1) | 319 (3.2) | <.001 |
| 60-69 | 1944 (8.9) | 1206 (10.1) | 738 (7.3) | <.001 |
| 70-79 | 3856 (17.6) | 2241 (18.8) | 1615 (16.1) | <.001 |
| 80+ | 14,822 (67.5) | 7583 (63.7) | 7239 (72.0) | <.001 |
|
|
||||
| Dementia | 5243 (23.9) | 2565 (21.5) | 2678 (26.7) | <.001 |
| Cancer | 7569 (34.5) | 4948 (41.5) | 2621 (26.1) | <0.001 |
| Cardiac | 1879 (8.6) | 898 (7.5) | 981 (9.8) | <.001 |
| Respiratory | 1456 (6.6) | 717 (6.0) | 739 (7.4) | <.001 |
| aOther than above | 5814 (26.5) | 2785 (23.4) | 3029 (30.1) | .362 |
| Missing | 1 (0.0) | 1 (0.0) | ||
|
|
||||
| 0-1 | 838 (3.8) | 396 (3.3) | 442 (4.4) | .011 |
| 2 | 2284 (10.4) | 1190 (10.0) | 1094 (10.9) | .030 |
| 3 | 6320 (28.8) | 3354 (28.2) | 2966 (29.5) | .026 |
| 4 | 12,512 (57.0) | 6971 (58.5) | 5541 (55.1) | <.001 |
| Missing | 8 (0.0) | 2 (0.0) | ||
|
|
||||
| 3218 (14.7) | 1817 (15.3) | 1401(13.9) | .006 | |
| 2 | 4148 (18.9) | 2278 (19.1) | 1870 (18.6) | .336 |
| 3 | 4574 (20.8) | 2372 (19.9) | 2202 (21.9) | <.001 |
| 4 | 6286 (28.6) | 3481 (29.2) | 2805 (27.9) | .032 |
| (Most deprived) 5 | 3210 (14.6) | 1681 (14.1) | 1529 (15.2) | .020 |
| Missing | 526 (2.4) | 284 (2.4) | 242 (2.4) | .885 |
|
|
||||
| Aged >18 and consented to record creation | 13,189 (60.1) | 7483 (62.8) | 5706 (56.8) | <.001 |
| Aged >18 but lacks capacity; BI/LPA decision | 8773 (39.9) | 4430 (37.2) | 4343 (43.2) | <.001 |
|
|
||||
| No | 6768 (30.8) | 8594 (72.1) | 3449 (34.3) | <.001 |
| Yes | 15,194 (69.2) | 3319 (27.9) | 6600 (65.7) | <.001 |
|
|
||||
|
|
||||
| For CPR | 3183 (14.5) | 1323 (11.1) | 1090 (10.9) | .637 |
| DNACPR | 18,779 (85.5) | 10,590 (88.9) | 8959 (89.2) | .637 |
|
|
||||
| No | 7194 (32.8) | 3532 (29.7) | 3662 (36.4) | <.001 |
| Yes | 14,768 (67.2) | 8381 (70.4) | 6387 (63.6) | <.001 |
|
|
||||
| No | 5645 (25.7) | 2722 (22.9) | 2923 (29.1) | <.001 |
| Yes | 16,317 (74.3) | 9191 (77.2) | 7126 (70.9) | <.001 |
|
|
||||
| Full sctive treatment inc. CPR | 2484 (11.3) | 1006 (8.4) | 1478 (14.7) | <.001 |
| Full active treatment but not CPR | 1457 (6.6) | 674 (5.7) | 783 (7.8) | <.001 |
| Treatment of any reversible conditions inc. hospital admission, | 5138 (23.4) | 2718 (22.8) | 2420 (24.1) | .028 |
| Treatment of any reversible conditions but only in the home/hospice setting | 7655 (34.9) | 4499 (37.8) | 3156 (31.4) | <.001 |
| Symptomatic treatment only | 4133 (18.8) | 2472 (20.8) | 1661 (16.5) | <.001 |
| Other | 76 (0.4) | 525 (4.4) | 494 (4.9) | .074 |
| Missing | 1019 (4.6) | 76 (0.4) | 57 (0.6) | <.001 |
|
|
||||
| Home | 12,524 (57.0) | 7460 (62.6) | 5064 (50.4) | <.001 |
| Care home | 6366 (29.0) | 3103 (26.0) | 3263 (32.5) | <.001 |
| Hospice | 197 (0.9) | 124 (1.0) | 73 (0.7) | .015 |
| Hospital | 741 (3.4) | 295 (2.5) | 446 (4.4) | <.001 |
| Other | 134 (0.6) | 62 (0.5) | 72 (0.7) | .058 |
| Not recorded | 2000 (9.1) | 869 (7.3) | 1131 (11.3) | .387 |
|
|
||||
| Home | 9907 (45.1) | 5955 (50.0) | 3952 (39.3) | <.001 |
| Care home | 5844 (26.6) | 2895 (24.3) | 2949 (29.4) | <.001 |
| Hospice | 1180 (5.4) | 791 (6.6) | 389 (3.9) | <.001 |
| Hospital | 454 (2.1) | 198 (1.7) | 256 (2.6) | <.001 |
| Other | 285 (1.3) | 132 (1.1) | 153 (1.5) | .009 |
| Not recoded | 4292 (19.5) | 1942 (16.3) | 2350 (23.4) | <.001 |
|
|
||||
| No | 4332 (19.7) | 2231 (18.7) | 2103 (20.9) | <.001 |
| Yes | 17,630 (80.3) | 9682 (81.3) | 7946 (79.1) | <.001 |
|
|
||||
|
|
||||
| Median (IQR) | 2 (0-5) | 3 (0-6) | 1 (0-3) | |
| Range | 0-112 | 0-112 | 0-40 | |
| At least 2 nonurgent care views | 11,655 (53.1) | 7621 (64.0) | 4034 (40.2) | <.001 |
|
|
||||
| Median (IQR) | 0 (0-1) | 0 (0-1) | 0 (0-1) | |
| Range | 0-23 | 0-22 | 0-23 | |
|
|
13,096 (59.6) | 7001 (58.8) | 6095 (60.7) | <.001 |
|
|
8866 (40.4) | 4912 (41.2) | 3954 (39.4) | <.001 |
Abbreviations: IMD, index of multiple deprivation; IQR, interquartile range; COVID-19, coronavirus disease 2019; WHO, World Health Organization; IMD, index of multiple deprivation.
Other diagnoses include neurological, gastro-intestine, renal, endocrine, haematological, mental health, musculoskeletal, vascular, COVID-19, “Other.”
Variables with significant missing data: ethnicity, capacity to make and communicate resuscitation decisions, receipt of care, family support.
Appendix III: Table of Forced-Multivariate Logistic Regression Models
| Forced Multivariate Logistic Regression Modelling for Dying in Preferred Place of Death | ||
|---|---|---|
|
|
|
|
|
|
||
| 1.00 (0.88 – 1.12) | 0.901 | |
| 0.83 (0.75 – 0.92) | <0.001 | |
| 1.14 (0.96 – 1.36) | 0.066 | |
| Cardiac | 0.69 (0.55 – 0.85) | 0.004 |
| Respiratory | 0.79 (0.62 – 1.01) | 0.158 |
| Other than above | 0.78 (0.66 – 0.92) | <0.001 |
| 1.05 (0.76 – 1.45) | 0.744 | |
| 3 | 1.05 (0.78 – 1.43) | 0.731 |
| 4 | 1.65 (1.22 – 2.24) | 0.002 |
| 0.88 (0.73 – 1.05) | 0.150 | |
| 3 | 0.77 (0.65 – 0.92) | 0.004 |
| 2 | 0.76 (0.65 – 0.90) | 0.002 |
| (Most Deprived) 1 | 0.65 (0.54 – 0.79) | <0.001 |
| 1.08 (0.93 – 1.25) | 0.298 | |
| 1.55 (1.23 – 1.95) | 0.001 | |
| 0.77 (0.61 – 0.96) | 0.018 | |
|
|
||
| 1.41 (0.96 – 2.07) | 0.091 | |
| Hospice | 0.90 (0.65 – 1.24) | 0.511 |
| Hospital | 1.66 (1.12 – 2.45) | 0.012 |
| Other | 0.05 (0.02 – 0.11) | <0.001 |
| 1.36 (0.92 – 2.01) | 0.125 | |
| Hospice | 0.65 (0.40 – 1.05) | 0.080 |
| Hospital | 0.99 (0.68 – 1.43) | 0.940 |
| Other | 0.97 (0.39 – 2.43) | 0.928 |
| Not recorded | 1.14 (0.60 – 2.17) | 0.715 |
| 1.61 (1.32 – 1.97) | <0.001 | |
| Symptomatic Treatment | 3.55 (2.78 – 4.53) | <0.001 |
| Other than the above | 1.46 (1.09 – 1.97) | 0.013 |
| 2.33 (1.75 – 3.10) | <0.001 | |
| 1.53 (1.34 – 1.74) | <0.001 | |
|
|
||
| 1.27 (1.13 – 1.43) | <0.001 | |
| 0.74 (0.66 – 0.82) | <0.001 | |
|
|
||
| 0.83 (0.69 – 0.99) | 0.037 | |
| London 3 | 0.95 (0.81 – 1.13) | 0.582 |
| London 4 | 0.72 (0.60 – 0.85) | <0.001 |
| London 5 | 0.68 (0.57 – 0.81) | <0.001 |
| Other | 0.73 (0.24 – 2.21) | 0.584 |
|
|
1.00 (1.00 – 1.00) | 0.018 |
| 1.06 (0.92 – 1.22) | 0.425 | |
| 1st Wave | 1.09 (0.97 – 1.23) | 0.129 |
| Forced Multivariate Logistic Regression Modelling for Dying in Hospital | ||
|
|
|
|
|
|
||
| 1.23 (1.11 – 1.36) | 0.008 | |
| 0.85 (0.75 – 0.96) | <0.001 | |
| 0.64 (0.52 – 0.78) | <0.001 | |
| Cardiac | 1.79 (1.43 – 2.24) | <0.001 |
| Respiratory | 1.45 (1.14 – 1.85) | 0.003 |
| Other than above | 1.42 (1.20 – 1.68) | <0.001 |
| 0.89 (0.67 – 1.17) | 0.394 | |
| 3 | 0.94 (0.73 – 1.22) | 0.644 |
| 4 | 0.59 (0.45 – 0.77) | <0.001 |
| 0.91 (0.76 – 1.09) | 0.282 | |
| 3 | 1.21 (1.02 – 1.45) | 0.031 |
| 2 | 1.19 (1.01 – 1.41) | 0.040 |
| (Most Deprived) 1 | 1.34 (1.11 – 1.62) | 0.003 |
| 1.03 (0.89 – 1.20) | 0.688 | |
| 0.78 (0.68 – 0.90) | 0.001 | |
|
|
||
| 0.60 (0.49 – 0.74) | <0.001 | |
| 0.66 (0.59 – 0.75) | <0.001 | |
| 0.51 (0.43 – 0.61) | <0.001 | |
| Symptomatic Treatment | 0.23 (0.18 – 0.29) | <0.001 |
| Other than the above | 0.52 (0.41 – 0.66) | <0.001 |
| 1.96 (1.60 – 2.38) | <0.001 | |
| Not Recorded | 1.72 (1.47 – 2.00) | <0.001 |
| 1.88 (1.42 – 2.50) | <0.001 | |
| Not Recorded | 1.00 (0.81 – 1.23) | 0.990 |
|
|
||
| 0.75 (0.67 – 0.84) | <0.001 | |
| 1.44 (1.30 – 1.61) | <0.001 | |
|
|
||
| 1.25 (1.04 – 1.50) | 0.017 | |
| London 3 | 1.02 (0.86 – 1.20) | 0.834 |
| London 4 | 0.92 (0.75 – 1.12) | 0.410 |
| London 5 | 1.68 (1.41 – 2.00) | <0.001 |
| Other | 1.37 (0.46 – 4.03) | 0.570 |
|
|
1.04 (1.00 – 1.08) | 0.026 |
| 0.96 (0.83 – 1.10) | 0.546 | |
| 1st Wave | 0.91 (0.81 – 1.02) | 0.113 |
Acknowledgments
The authors wish to thank the Coordinate My Care information team, for extracting the data for analysis. We would like to thank Lisa Scerri and the Royal Marsden Biomedical Research Centre for their support of this study and the use of the Biomedical Research Informatics Digital Environment, a Trusted Research Environment.
Authors’ Contributorship
PM, JD, MO, JRo, AB, MP, and JRi designed the study. JD led the PPI work. PM analyzed and interpreted the data with statistical support provided by CB and JD. JD, JRo, AB, MP, KES, MO, JK, and JRi provided support in interpreting the data. PM drafted the manuscript. All authors provided have revised the article critically for clarity and intellectual content provided revisions and have approved this version for submission.
Data Management and Sharing
Anonymized data from this study are stored on the Biomedical Research Informatics Digital Environment, a Trusted Research Environment and informatics platform at The Royal Marsden Biomedical Research Centre. Data sharing requests and access to the protocol and supplementary information would be available on reasonable request after the completion of existing studies and whenever legally and ethically possible. Data access requests should be directed to Dr Joanne Droney joanne.droney@rmh.nhs.uk. Once approved, projects intending to use the data will be reviewed and authorized by The Royal Marsden Committee for Clinical Research Committee and access will be provided via the Trusted Research Environment.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest concerning the research, authorship, and/or publication of this article.
Ethical Approval
The study used anonymized data collected in routine clinical practice and did not require ethical approval. Patients consent to the use of their anonymized data for research when creating a Coordinate My Care record. The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project represents independent research supported by the National Institute for Health and Care Research (NIHR) Biomedical Research Centre at The Royal Marsden NHS Foundation Trust and the Institute of Cancer Research, London. PPI was funded through a PPI NIHR BRC grant. The researcher's salary was funded through the Mary Hambro Research Fellowship Fund, part of the Coordinate My Care Charity Research Fund. The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care.
Correction (August 2023):
Appendix III was incorrectly showing wrong table. This has been corrected now.