Abstract
Early mobilisation following elective total hip arthroplasty (THA) facilitates quicker rehabilitation, and reduces complications and hospital length of stay. Reasons for delayed mobilisation are multifactorial, but the most common cause is orthostatic intolerance. Midodrine, an oral alpha-1 agonist, is used off-label for perioperative hypotension. However, there are few randomised trials assessing its use in the perioperative setting to improve patient outcomes. The aim of the study was to determine whether midodrine improves early mobilisation following primary THA, and whether this relates to reduced orthostatic intolerance. This prospective, triple-blinded, multicentre study involved 42 patients randomised to either placebo or 20 mg midodrine, 2 h before physiotherapy, on Day 1 postoperatively. The inclusion criteria were adults undergoing elective unilateral THA under spinal anaesthesia. The primary endpoint was the ability to walk 5 m with physiotherapists. Secondary endpoints included the incidence of orthostatic intolerance and hypotension. A preplanned interim analysis showed no statistical difference in ability to mobilise 5 m (78.26% vs 78.95%, P = 1.0). There was no statistically significant difference in the incidence of orthostatic intolerance between the groups 17.4% vs 31.6% (P = 0.45). Pre-emptive use of midodrine did not improve patient mobilisation the morning after elective primary THA and had no significant effect on the incidence of orthostatic hypotension.
Introduction
Total hip arthroplasty (THA) surgery is a common operation with approximately 50,000 cases annually in Australia, and over 775,589 primary hip arthroplasties performed between 2012 and 2020 in America.1,2 This number is expected to rise and improvements in perioperative care, including reducing the length of stay, will be required to manage the economic burden associated with this episode of health care. Early independent mobilisation is an important patient-centred outcome and is integral to facilitating earlier discharge from hospital.
There are several reasons why patients may fail to mobilise, but orthostatic intolerance is a common and potentially preventable cause. 3 Orthostatic intolerance is defined as symptoms of dizziness, nausea, blurred vision or syncope when moving from a lying to a sitting or standing position. Orthostatic hypotension is defined as a drop in systolic blood pressure (SBP) of 20 mmHg or a diastolic blood pressure (DBP) drop of 10 mmHg. 4 Orthostatic intolerance may occur without demonstrable changes in blood pressure and similarly not every episode of orthostatic hypotension induces symptoms of intolerance. Following elective hip arthroplasty, the incidence of postoperative orthostatic intolerance ranges from 19% to 52%.5 –8 In our institution, an audit of 300 patients undergoing arthroplasty surgery found that up to 45% experienced hypotensive episodes in the first 24 h after surgery.
Perioperative blood pressure and orthostatic intolerance
Perioperative blood pressure is a complex, broad topic with considerable interest in its detection, prevention and management, to reduce potential morbidity and improve outcomes.9 –16 Pathological causes of postoperative hypotension may include serious conditions such as active bleeding, pulmonary embolism, myocardial infarction, and sepsis. Other causes of perioperative hypotension may have several contributory factors. These include preload reduction due to hypovolaemia and intraoperative blood loss; impaired contractility due to myocardial disease; and afterload reductions secondary to surgical trauma, neuraxial techniques, anaesthesia, opioids and antihypertensive medications.17 –22 While acknowledging these contributory factors, in its simplest form, orthostatic intolerance is a form of autonomic dysregulation with failure of adequate baroreceptor responses to standing, resulting in inability to mobilise. 6 The exact pathology is not clear, however female sex, high pain levels during mobilisation, and use of gabapentinoids have been reported as risk factors. 5 Current treatment strategies such as bed rest and intravenous fluid administration are limited in efficacy. 23 While intravenous vasopressors are effective, they can generally only be administered in high dependency areas, which is impractical.
Midodrine
A potentially simpler and cheaper alternative treatment to orthostatic intolerance is oral midodrine. Midodrine, an oral α1-agonist, can increase venous return and improve systemic vascular resistance. 24 It has been approved ‘for the treatment of severe symptomatic orthostatic hypotension due to autonomic dysfunction when exacerbating factors have been addressed and other forms of treatment remain inadequate’. 25 Midodrine use has previously been widely reported in the non-operative setting with some promising results, but these have largely been observational studies.26 –41 It has also been evaluated in the intensive care setting as an adjunct for weaning vasopressors with some initial enthusiasm in observational studies, but unsubstantiated in randomised trials.42 –45 There are limited data in the operative setting with reports in transplant surgery, spinal surgery and carotid artery stenting yielding mixed results.46 –52
In a pilot study of 20 patients undergoing hip and knee arthroplasty, variable dose midodrine given for the first 72 h after surgery was found to significantly increase SBP response. 53 However, in a larger randomised trial of 120 hip arthroplasty cases, a single 5-mg dose of midodrine administered 1 h prior to mobilisation did not significantly reduce the prevalence of orthostatic hypotension. 54 Further dose finding studies were recommended.
Its side effect profile has been considered excellent with the main significant side effect being induced hypertension. 55 Furthermore, it does not cross the blood–brain barrier, hence has no central nervous system side effects. Other minor side effects that have been reported include the urge to urinate, skin tingling and pruritus. 56
The aim of this study was to assess whether midodrine would successfully improve mobilisation following hip arthroplasty, predominantly by improving orthostatic tolerance.
Materials and methods
Study design
The study was a prospective, multicentre, randomised, controlled, triple-blinded (participants, experimenters and study investigators), superiority clinical trial conducted at two sites (a quaternary teaching hospital and a district general hospital). The trial was registered with the Australian New Zealand Clinical Trials Registry (Study ID: ACTRN12620001267943). The trial was partially funded via a grant provided by the Research and Advisory Committee of the Sir Charles Gairdner and Osborne Park Health Care Group. The study was approved by the Sir Charles Gairdner and Osborne Park Health Care Group Human Research Ethics Committee (RGS0000003096). Written informed consent was obtained from all participants enrolled into the study. There were no changes to the trial design following commencement of the study.
Adult patients (>18 years of age) undergoing an elective unilateral THA were identified from theatre booking lists (Theatre Management System) and deemed eligible for the study recruitment between January 2021 and February 2022. However, patients who met the criteria in Table 1 were excluded.
Exclusion criteria.
NYHA: New York Heart Association.
Anaesthesia application
All patients received standard monitoring. Large bore intravenous (IV) access was secured, and compound sodium lactate commenced with all patients receiving 1–2 L intraoperatively.
Sedation with 1–2 mg midazolam IV was given prior to administration of 2.0–3.0 ml of intrathecal isobaric bupivacaine 0.5%. Additional IV medications included cephazolin 2 g, dexamethasone 8 mg, parecoxib 40 mg, granisetron 1 mg, and tranexamic acid 1 g. Sedation was administered with a propofol infusion titrated to a depth of sedation at the discretion of the anaesthetist (no clonidine or dexmedetomidine was administered). A laryngeal mask was inserted if a general anaesthetic was required with supplemental 1.0–1.5 micrograms/kg IV fentanyl on induction. An indwelling urinary catheter was inserted if indicated according to local policy (history of incontinence, previous history of postoperative urinary retention, or a bladder scan volume postoperatively demonstrating >500 ml with inability to void). Any intraoperative hypotension (defined as SBP <100 mmHg) was managed clinically by the anaesthetist using IV Metaraminol 0.5 mg bolus and/or continuous infusion (but not intramuscular or subcutaneous ephedrine or metaraminol or oral midodrine). Orthopaedic surgeons infiltrated 100 ml of 0.2% periarticular ropivicaine. Postoperatively, all patients were prescribed a further 1 g of tranexamic acid after 6 h, and 1 L IV compound sodium lactate fluid. In the post anaesthesia care unit, a ‘rescue’ dose of midodrine was permitted for refractory symptomatic hypotension (SBP <100 mmHg with dizziness/drowsiness, blurred vision or nausea) using a pre-existing local hospital protocol.
Group Allocation and randomisation
Randomisation occurred using a computer-generated allocation sequence in a 1:1 method by an independent investigational pharmacist not involved in the study. On the day of surgery, the experimenters allocated patients to a trial medicine stored in a container, which was either the active medication, midodrine 20 mg (in 4 × 5 mg tablets), or placebo (also in 4 tablets). The trial medicines were not identifiable and stored in identical packaging. The study medicine was prescribed on the patient’s prescription chart and administered by a ward nurse 2 h before physiotherapy on Day 1 postoperatively. If the supine SBP ≥160 mmHg at the time of administration, the study medicine was withheld and disposed of.
Baseline measurements
Preoperative demographics included age, height, weight, American Society of Anesthesiologists Physical Status Classification, co-morbidities, and medications including antihypertensives and diuretics. Additional data included preoperative blood pressure and heart rate, and pathology results including haemoglobin, creatinine, ferritin and troponin.
Evaluation of outcome measures
The primary outcome was the ability to successfully mobilise at least 5 m on the first attempt during the morning of Day 1. The process was as follows:
supine blood pressure was measured by a physiotherapist using an automated sphygmomanometer 15 min pre-mobilisation; following this, the patient sat on the edge of the bed and did marching on the spot exercises; if tolerated, the patient proceeded to stand upright (with the assistance of at least one person and a frame as needed); a standing blood pressure was measured after 1 and 3 min; if standing was tolerated, then the patient proceeded to attempt walking (with or without aids); if unable to mobilise, a reason was documented such as pain, orthostatic intolerance or hypotension.
The secondary outcomes included:
ability to mobilise at least 10 m on Day 2 postoperatively; incidence of orthostatic intolerance at the first mobilisation (defined as an inability to mobilise caused by the onset of one or all of the following symptoms: dizziness, blurred vision, nausea or vomiting, syncope after sitting or standing from a supine position regardless of the blood pressure); incidence of orthostatic hypotension postoperatively (defined as a drop in SBP of 20 mmHg and/or a DBP drop of 10 mmHg as measured using an automated sphygmomanometer); supine hypotension on Day 1 following surgery (defined as a recorded SBP <100 mmHg at any time during normal observations); changes in blood pressure post study drug administration at 1, 2, 3 and 4 h; hospital length of stay (days as an inpatient); potential side effects of midodrine in the 24-h period following study drug administration: supine hypertension (SBP >180 mmHg, or DBP >110 mmHg), severe bradycardia (heart rate <40/min), urinary retention requiring catheterisation, severe pruritus; this was monitored by an independent safety committee; pain score before and after physiotherapy on a numeric rating scale from 0 (no pain) to 10 (worst pain).
Sample size
The sample size estimation was calculated based on published data and a local pilot study demonstrating a baseline of 40% of patients that failed to mobilise on the first attempt the day after surgery.5 –8 A reduction in the failure rate to 10% was accepted as clinically significant.
Assuming an alpha error of 0.05 and a beta error 0.2, a total of 76 patients was required. Considering an approximate 10% loss to follow-up, the investigators decided 84 patients were required in total. At the halfway point of the study, a planned interim analysis was performed to recalculate the power and ensure group sizes were adequate. The interim analysis was performed by an independent statistician blinded to the treatment allocation.
Statistical analysis
The statistical analysis was conducted using Python Jupyter Notebook and utilising the SciPy statistics package. Data were analysed as intention-to-treat. Dichotomous outcomes were analysed using the chi-square test (e.g. successful mobilisation postoperatively, incidence of orthostatic intolerance and orthostatic hypotension). The Mann–Whitney U test was used for integer data, such as hospital length of stay. Continuous data were analysed using the Student’s t-test. Pain scores were reported as mean (interquartile range, IQR). A P-value <0.05 was accepted as statistically significant.
Results
Between 5 January 2021 and 24 February 2022, we recruited and randomised 42 patients to the study (Figure 1). Twenty-three patients were allocated to the placebo group and 19 to the midodrine group. Patient characteristics and intraoperative management were similar between groups (Tables 2 and 3).
CONSORT diagram for the trial. Randomisation, allocation and follow-up for each group.
Baseline characteristics.
ASA PS: American Society of Anaesthesiologists Physical Classification Status; SBP: systolic blood pressure; HR: heart rate; BMI: body mass index; Pre-op: preoperative; BPM: beats per minute; SD: standard deviation; IQR: interquartile range.
Data are mean (SD) or n (%), other than ASA PS reported as median (IQR).
Intraoperative data.
Intra-op: intraoperative; IV: intravenous; SBP: systolic blood pressure; SD: standard deviation; IQR: interquartile range; GA: general anaesthesia.
Data are mean (SD) or n (%) or median (IQR) for spinal fentanyl dose.
In regard to the primary outcome, 5 (21.7%) in the placebo group failed to successfully mobilise 5 m on Day 1 postoperatively, compared to 4 (21.1%) in the midodrine group (P = 1.00). In all instances the reason for failed mobilisation was orthostatic hypotension or symptoms of orthostatic intolerance. There was no difference between the groups regarding ability to successfully mobilise >10 m on Day 2 (placebo 91.3% vs midodrine 94.7%, P = 1.00). The incidence of orthostatic intolerance was 17.4% vs 31.6% (P = 0.45) and orthostatic hypotension 37.5% vs 18.2% (P = 0.19) in the placebo and midodrine groups respectively. Supine hypotension occurred in 8.7% (n = 2) in the placebo group compared to 21.1% (n = 4) in the midodrine group (P = 0.38). One patient in each group required a dose of rescue midodrine in recovery.
The changes in blood pressure in the 4 h following administration of the study medication were not significant (P = 0.24; see Figure 2). Additional data on intraoperative management, postoperative patient and haemodynamic data and length of stay are presented in Table 4 and Figure 2.
Changes in blood pressure in the first 4 h following study medication administration. Systolic and diastolic parameters for each group are displayed on the y-axis with time on the x-axis. The symbols refer to mean and error bars within 2 standard deviations.
Postoperative (Day 1) patient and haemodynamic data.
OME: oral morphine equivalent; PONV: postoperative nausea and vomiting; LOS: length of stay; SD: standard deviation; IQR: interquartile range.
Data are mean (SD) or n (%) or median (IQR) for pain scores.
After the interim analysis, the data and safety monitoring board recommended the trial be stopped early. The reasons for this were (1) a P-value for efficacy of 1.00 and (2) a low probability of demonstrating a statistical difference, even with adjustment of the sample size to account for the lower-than-expected event rate. The principal investigator agreed to stop recruitment and proceed to the final data analysis.
Discussion
We conducted a triple-blinded, multicentre, randomised controlled trial to determine whether midodrine administration could reduce the incidence of orthostatic hypotension or intolerance and improve mobilisation following THA. Our pragmatic study showed that administration of 20 mg midodrine prior to physiotherapy did not improve the rate of successful mobilisation on the day following surgery. Furthermore, there was no measurable difference in blood pressure between the two groups. A preplanned interim analysis was conducted by an independent statistician blinded to the treatment allocation and led to the decision to cease the trial early due to futility.
There are several potential reasons why midodrine was ineffective. Firstly, the dose may have been inadequate or the timing incorrect. Prior studies have demonstrated a linear dose–response relationship suggesting higher doses are more effective than lower doses.39,57 In a large randomised trial of 120 hip arthroplasty cases, a 5-mg dose of midodrine administered 1 h prior to mobilisation did not significantly reduce the prevalence of orthostatic hypotension. 54 In a smaller pilot study, incremental doses ranged from 2.5 to 10 mg given three times per day during the first 72 h. 53 Lower does were found to be ineffective but a dose of 10 mg was found to achieve a significant SBP response 2 h after administration (mean 14 mmHg). In patients who received higher dose midodrine, adverse mobilisation events appeared less common. They recommended further randomised trials using 10 mg 2 h prior to mobilisation to assess efficacy. Our dose of 20 mg was higher than previously used but may still have been inadequate.
In terms of timing, oral administration of midodrine typically results in almost complete absorption within 1 h with 93% bioavailability. 58 The timing of 2 h was chosen due to previous observation of SBP rise peaking at 2-h post midodrine administration. 53 However, it is possible that the timing was incorrect with either slow onset due to variable gastric absorption because of opioid effects or rapid clearance—the terminal elimination life is 3–4 h.
Interestingly, we did not observe any measurable rise in SBP or any difference between midodrine and placebo. Potential explanations include insufficient frequency of measurements or measurement errors. An automated sphygmomanometer was used in a standardised fashion in both groups to reduce error. An arterial line with continuous readings would undoubtedly improve precision, but it was impractical in this setting. Preload effects clearly need to be considered but there were no measurable differences in fluid administration, operative duration or haemoglobin concentration between the groups. As an alpha-1 agonist, midodrine would be expected to increase afterload and SBP, however, the absence of a beta cardiac stimulant effect may result in a reduction in cardiac output. In addition, patients may experience a reflex reduction in heart rate. These effects could potentially be offset by the addition of ephedrine, which may be grounds for future trials.
Dynamic pain control is also required to aid mobilisation. We used a consistent prescribing protocol supervised by an acute pain team. All patients were prescribed regular paracetamol and celecoxib, sustained release tapentadol twice daily (or oxycodone hydrochloride/naloxone hydrochloride anhydrous combined tablet if there was an adverse reaction to tapentadol) and a PRN oral opioid at the discretion of the anaesthetist – this was normally either oxycodone, buprenorphine or hydromorphone.
Additional factors that may affect successful mobilisation that are difficult to measure include patient education and motivation, institutional culture and practices, physiotherapy resources, levels of staffing, and availability of adjuncts such as appropriate footwear and walking aids. One of the sites had a single physiotherapist mobilising patients postoperatively; however, the other site had a team of physiotherapists of variable seniority and experience. We also acknowledge that a placebo effect could affect patients’ expectation of their ability to mobilise. However, there is no clear physiological explanation of how a placebo effect could offset or reverse the anticipated vasodilatory effects underlying the aetiology of orthostatic intolerance.
Our study had several limitations. Firstly, it was underpowered. We had anticipated that up to 40% of patients in the control group would fail their first mobilisation attempt, whereas the actual rate was 21%. The baseline rate in the literature is highly variable ranging from 12% to 70% in the general surgical population, and 19% to 42% following hip arthroplasty.2 –8 In our institution, an audit of 300 patients undergoing arthroplasty surgery found that up to 45% experienced hypotensive episodes in the first 24 h after surgery. Furthermore, a locally conducted randomised trial assessing analgesic efficacy of an erector spinae plane block (ESPB) in hip arthroplasty by the same authors reported Day 1 mobilising failure rates of 42%, 5 which contributed to our expectation of a higher incidence of orthostatic intolerance. Of note, gabapentinoids were not used in this study but had been used in the ESPB trial and may have contributed to the higher incidence of orthostatic intolerance.
A second limitation was that the timing of analgesia administration in relation to episodes of mobilisation was not recorded. As such we are unable to comment on the effect of analgesic administrative timing. Irrespective of the exact timing of analgesic administration, there was no difference in either overall oral morphine equivalent or pain scores during physiotherapy between the groups and no patient failed to mobilise because of inadequate analgesia.
Independent statistical calculation at the interim analysis demonstrated that higher recruitment rates would have been futile and as such the trial was ceased early. Post hoc contingency analysis with recalculation of the sample size to 152 patients and a control group event rate of 21% indicated the data would be statistically insignificant (P-value 0.23) even if all patients in the midodrine group successfully mobilised. Future trials in this area would benefit from pilot studies of baseline mobilisation rates in their anticipated study groups to guide powering more accurately.
While we are unable to confirm or refute midodrine’s ability to prevent failed mobilisation following THA, this study suggests the treatment effect, if any, may be of a small and clinically insignificant magnitude.
Footnotes
Author Contribution(s)
Acknowledgements
We would like to thank the Research and Advisory Committee of the Sir Charles Gairdner and Osborne Park Health Care Group for funding the research. We would also like to thank Dr Charlotte Boyle (Sir Charles Gairdner Hospital) for her assistance in reviewing, editing and in preparing the final submission of this paper.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author on request.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially funded by the Research and Advisory Committee of the Sir Charles Gairdner and Osborne Park Health Care Group (grant no. 2019-20/013).
