Frailty

Frailty is a multidimensional syndrome that occurs when multiple small physiological deficits accumulate gradually, eventually resulting in a loss of physiological reserve and adaptability, putting a patient at a higher risk of adverse outcomes.1 The prevalence of frailty before cardiac surgery varies among studies but generally ranges from 20% to 50%.2 In older adults undergoing aortic valve replacement, the prevalence of frailty ranged from 26% to 68%, depending on the frailty instruments used and the type of surgical approach used in a multicentre study of 1020 patients from 2012 to 2016.3 In another study, pre-frailty (Clinical Frailty Scale [CFS] score of 4)1 was present in 60% of patients undergoing elective coronary artery bypass grafting with or without valve replacement.4 These results suggest that frailty is becoming a progressively important public health problem. Despite increasing recognition that frail individuals are more vulnerable to sudden changes in health status (stressors),5 6 most clinicians can identify frailty before surgery but often ignore it.7

Preoperative frailty is associated with poorer clinical outcomes and higher healthcare costs in patients undergoing cardiac surgery. Even adjusting for routine risk scoring systems in cardiac surgery, frailty was associated with a twofold to sixfold increased risk of major adverse cardiac and cerebrovascular events (MACCEs), prolonged hospital stay, in-hospital mortality and 1-year mortality in two systematic reviews.8 9 In another study, frailty was also associated with a threefold to eightfold increased risk of postoperative delirium.10 Not surprisingly, the total cost of hospitalisation associated with frailty was high after adjusting for age, gender, predicted risk of mortality and type of cardiac procedure (adjusted additional cost in Canadian dollars: $21 245, 95% CI $12 418 to $30 073; p<0.001).11

Vulnerable (pre-frail) patients may also be at risk of poor patient outcomes after cardiac surgery. Compared with non-frail patients, pre-frail patients (CFS of 4) had a longer intensive care unit (ICU) stay (mean difference 2.0 days, 95% CI 1.7 to 2.3), a longer hospital stay (mean difference 3.0 days, 95% CI 1.8 to 4.2), a higher risk of postoperative stroke (HR 1.8, 95% CI 1.3 to 2.5) and a higher risk of in-hospital mortality (HR 1.8, 95% CI 1.5 to 2.3).4 In patients undergoing transcatheter aortic valve replacement, the cumulative short-term mortality and long-term mortality was higher in the pre-frail group than in the non-frail group (30 days: 1.8% vs 1.3%, 1 year: 8.6% vs 7.2%).12

Prehabilitation

Managing pre-frail to moderately frail patients may be an effective way to prevent, delay or even reverse frailty. Physical activity may be an important component of frailty management. In particular, exercise training has been shown to improve various aspects of physical function of the frail elderly (eg, muscle strength, body composition, mobility, functional status and fall prevention) in a recent systematic review of nine randomised controlled trials (RCTs) (n=1067).13 Tailored exercise training is hence expected to improve physical fitness and increase functional capacity so that patients are better prepared to withstand the consequences of the physical stress of surgery.14 In a systematic review, prehabilitation in 643 older adults before orthopaedic surgery had beneficial effects (effect size >0.2) in improving strength, flexibility, balance and speed in five of the seven RCTs.15 In addition to structured exercise training, many ‘surgical schools’ also incorporate other health promotion activities, such as smoking cessation, nutritional advice and psychological support to reduce preoperative stress.16 Preliminary evidence suggests that a brief cognitive behavioural intervention with activity and risk factor reduction goals are more than 90% likely to be cost-effective if the cost per quality adjusted life years was less than £30 000.17

Physical therapy before cardiac surgery, especially inspiratory muscle training and breathing exercises, reduced the risk of postoperative pneumonia (risk ratio [RR] 0.45, 95% CI 0.24 to 0.83), postoperative atelectasis (RR 0.52, 95% CI 0.32 to 0.87) and shortened the length of postoperative stay in the hospital (mean difference −3.21 days, 95% CI −5.73 to −0.69).14 However, the systematic review of eight RCTs (n=856)14 did not address the potential benefits of total body exercise or assess frailty in the included studies. Also, the external generalisability to our setting is questionable as three of the included studies (n=347) had high prevalence of patients with chronic obstructive pulmonary disease (>20%) compared with a substantially lower prevalence in Hong Kong, where a 5% prevalence of patients with a preoperative history of asthma or chronic obstructive pulmonary disease has been reported.18

The timing and duration of prehabilitation programmes before elective cardiac surgery appear to be associated with the risk of postoperative outcomes. In 117 patients undergoing cardiac surgery, randomly allocated to either light physical exercise (1 hour, two times per week) and mental stress reduction or standard care within 2 weeks of surgery, there was little or no benefit on improving perioperative quality of life, length of stay in the hospital (p=0.54) or risk of postoperative atrial fibrillation (p=0.71).19 In another pilot RCT (n=17), prehabilitation exercise and health promotion (medication use, stress, diet and cardiovascular risk factor management) for 1 hour two times per week for at least 4 weeks showed outcome effectiveness using the 6 min walk test as the primary outcome.20 Patients in the prehabilitation group were able to walk further by 136 m (95% CI 61 to 209) and increase gait speed by 27% before surgery, both clinically meaningful improvements.20 A larger RCT with 244 pre-frail to moderately frail participants from the same research group is ongoing to assess the effectiveness of prehabilitation on hospital length of stay.21 Moderate physical exercise (1 hour, two times per week) for 8 weeks in low-risk cardiac surgical patients was associated with shorter ICU stay (median 1.5 hours, 95% CI 0.2 to 4.5), shorter hospital stay (median 1.0 day, 95% CI 0 to 1.0) and improved quality of life up to 6 months after surgery.22 Notably, only one of these RCTs20 measured frailty levels by the 5 m gait speed test and reported changes in frailty levels between prehabilitation groups over time.

The effect of self-reported preoperative physical activity on patient outcomes after cardiac surgery was mixed in a recent systematic review of 11 cohort studies (n=5733), partly reflecting the unreliability of self-reported preoperative physical activity questionnaires.23 Taken together, these findings suggest that prehabilitation for cardiac surgical patients may be an effective intervention if given over 1–2 months before surgery with objective preoperative measures of physical activity. Whether prehabilitation works by improving frailty levels is not well established. This study attempts to address this knowledge gap. The main hypothesis is that prehabilitation improves the level of postoperative recovery and reduces the risk of adverse outcomes following cardiac surgery by improving frailty levels.

We plan to conduct the PREhabilitation for improving QUality of recovery after ELective cardiac Surgery trial. The primary objectives of the study are

1. To examine the effect of prehabilitation on quality of early to short-term recovery after elective cardiac surgery in pre-frail and frail patients.

2. To determine how much the effect of prehabilitation on quality of early to short-term recovery is associated with improved frailty levels.

A secondary objective of the study is to determine if prehabilitation in pre-frail and frail patients is a cost-effective intervention for elective cardiac surgery.

Trial design

The study design is a single-centre, single-blinded, two-group, parallel, superiority, stratified RCT of 164 adults undergoing general anaesthesia for elective primary cardiac surgery (coronary artery bypass graft, valve replacement or both). The study was developed according to the Consolidated Standards of Reporting Trials  statement,24 and reported with reference to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT)25 and its Patient-Reported Outcome Extension (SPIRIT-PRO)26 statements. The trial has been registered on the Chinese Clinical Trials Registry. An overview of the study design is provided in figure 1. Neither were the patients involved in commenting on the study design nor did they contribute to the editing of this document for readability or accuracy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Patient flow.

Study setting and population

The study will be conducted at the Prince of Wales Hospital, a university teaching hospital with 1650 beds in Hong Kong. All elective cardiac surgical patients are admitted to our 23-bed ICU for early postoperative care and monitoring with 1:1 nursing at all times, with an expectation of discharge from ICU to a high-dependency cardiac ward within 24 hours after surgery. Currently, 25–30 adults undergo elective coronary artery bypass and/or valvular surgery per month.

Randomisation and allocation concealment

The stratified randomisation will be a 1:1 allocation according to a computer-generated sequence, using WINPEPI software V.11.65, performed by the one of the investigators (AL) who is not involved in screening, patient recruitment, clinical care or data collection. We will stratify the randomisation into three surgical groups (coronary artery bypass graft with or without valve repair/replacement, isolated valve repair/replacement and transcatheter aortic valve implantation), which have different cardiopulmonary bypass times and recovery profiles, based on case-mix prevalence in order to balance the groups. The treatment allocation will be concealed in consecutively numbered sealed opaque envelopes.

Eligibility criteria

We will use inclusion and exclusion criteria similar to those used by others.21 The inclusion criteria are

1. Adults (no age restriction) undergoing elective primary isolated coronary artery bypass grafting, aortic valve repair/replacement, mitral valve repair/replacement or combined coronary artery bypass/valve procedures.

2. Pre-frail to moderately frail patients with a CFS of 4–6 at the time of accepting surgery at the outpatient cardiothoracic surgical clinic.

3. Patients with an estimated 8 or more weeks of surgical waiting list time. One of the coinvestigators (MJU) will ensure that the clinically appropriate listed surgery date is compatible with an optimal prehabilitation time of 6–10 weeks.

The exclusion criteria are

1. Patients with unstable or recently unstable cardiac syndrome (New York Heart Association Class IV, critical left main coronary disease, hospitalisation for arrhythmias, congestive heart failure or acute coronary syndrome before randomisation).

2. Patients with severe left ventricular obstructive disease (severe aortic or mitral stenosis and dynamic left ventricular outflow obstruction).

3. Redo cardiac surgery.

4. Contraindications for prehabilitation, such as those with cognitive deficits who are unable to comply with study procedures, physical limitations that would preclude rehabilitation and inability to regularly attend outpatient prehabilitation sessions, such as those patients who are severely frail (CFS 7–9).

Participant screening and frailty assessments

For the purposes of this study, we will use three frailty tools to encompass the various domains of the frailty syndrome since there is no ‘gold standard’ definition for frailty. These domains include physical activity, mobility, energy, strength, cognition, independence and nutritional status.3 First, the use of the CFS is becoming widespread in the cardiac surgery and ICU settings as it is a simple, quick and highly predictive semiquantitative tool that can be used by nongeriatricians.3 12 27 The CFS ranges from 1 (very fit) to 9 (terminally ill), with common cut-off categories classified as non-frail (scores 1–3), vulnerable/pre-frail (score 4) and frailty (scores 5–9).1 Second, the 5 m gait speed has been used as a screening tool for frailty in the Society of Thoracic Surgeons database to predict all-cause mortality and morbidity during the first 30 days after cardiac surgery.28 Finally, the Essential Frailty Toolset (EFT) is a score from 0 (least frail) to 5 (most frail) derived from the time taken for five chair rises, mini-mental state examination and  haemoglobin and serum albumin concentrations.3 The EFT test had marginally higher discrimination properties for 30-day mortality and worsening disability at 1 year, as well as for mortality at 1 year than the CFS (area under the receiver operating characteristic curve 0.78 vs 0.74).3 We have chosen these measures based on the need for benchmarking and good diagnostic accuracy properties.

A physiotherapist will first screen patients for their eligibility to participate in the study. Consecutive eligible patients with a scheduled operation date will be recruited into the study. The sealed opaque envelope will be opened after obtaining written informed consent. All patients will have a thorough evaluation of medical history, physical function (6 min walk test [6MWT] for cardiorespiratory fitness testing, 5 m gait speed test as one of the screening tool for frailty and 30 s chair stand test for lower limb muscular strength and endurance testing), recent cardiovascular tests and procedures, current medications (β-blockers, calcium channel blockers, ACE inhibitors/angiotensin receptor blockers, nitrates, aspirin and antiplatelets) and other identifiable cardiovascular risk factors.

Blinding

A research nurse, collecting the postoperative outcomes (via face-to-face interview and telephone follow-up), will be blinded to treatment allocation to reduce measurement bias. Due to the nature of the intervention and the requirements of informed consent, trial participants may not be blinded to the treatment allocated with certainty. The physiotherapist, cardiothoracic surgeons, ICU and ward staff will not be blinded to intervention allocation because they have access to the patient’s electronic record of prehabilitation visits using the Hospital Authority Clinical Management System.

Interventions

All patients will receive standardised surgical processes and perioperative care under existing protocols for preoperative patient education, standardised anaesthesia,29 postoperative ICU sedation, analgesia and weaning from mechanical ventilation, perioperative physiotherapy, and early mobilisation. Participants assigned to the control arm of the study (table 1) will be given current usual care without a structured exercise and preoperative health promotion/patient education programme (no prehabilitation).

In contrast, participants randomised to the intervention arm (table 1) will be given prehabilitation (two times per week, supervised by a physiotherapist) in addition to current usual care. Patients will undergo a structured 6–10 weeks of preoperative exercise training to optimise the physical and psychosocial fitness of patients with underlying frailty syndrome before surgery in a dedicated room with gymnasium equipment at the Day Surgery Centre, Prince of Wales Hospital. The result of the submaximal exercise test from the 6MWT will be used as an estimation of individual peak oxygen uptake and hence oxygen uptake reserve (VO₂R) for exercise prescription.30 Individual heart rate performance and subjective perceived exertion will be monitored continuously throughout exercise. Each patient’s prehabilitation programme will be individualised and symptom limited, in which exercise prescription and progression will be based on the results of the exercise test, individual health status, exercise performance and training response.

Based on the American College of Sport Medicine guidelines on exercise prescription,30 the exercise protocol will include 5–10 min warm-up and cool-down activities, 20–60 min of aerobic exercises (with training intensity between 40% and 80% of VO₂R), and resistance training using cuff weight, which began with very light-to-light intensity (40%–50% of one repetition maximum), for 10–15 repetitions per major muscle groups of upper and lower limbs with one to three sets, depending on individual tolerance and performance. A home exercise programme will be encouraged with individualised exercise prescription, monitoring advice and recording sheet so that patients can perform 3–5 days of aerobic exercise per week in total (in the form of walking/running and stepping mainly) and 2–3 days of resistance exercise (same exercises prescribed during supervised sessions). Additional education on breathing techniques and daily activities will be given by a physiotherapist supervising the prehabilitation programme for all participants for consistency. Verbal and written advice on nutrition, smoking cessation and positive psychology support will be given to patients during the prehabilitation programme. Prehabilitation will be offered to those who have completed the whole series of training in order to improve the adherence and compliance at no cost.

Data collection

The physiotherapist will collect frailty measures (CFS, 5 m gait speed and EFT) before prehabilitation begins in both groups (table 2) and will collect patient’s compliance with attendance at prehabilitation visits in order to examine any dose–response effects. The research nurse will collect the following demographic data: age, gender, education level, American Society of Anesthesiologists’ Physical Status, predicted mortality using the logistic European System for Cardiac Operative Risk Evaluation (EuroScore), details of surgical procedures, duration of anaesthesia, duration of cardiopulmonary bypass time, ICU admission severity of illness score (APACHE III), duration of mechanical ventilation, ICU length of stay, duration of the hospital stay, discharge destination (home, hospital or nursing aged-care facility), 30-day mortality and 90-day mortality status from the patient’s medical record. To reduce detection bias, the research nurse, a blinded outcome assessor, will perform frailty tests at hospital admission and collect CFS scores after surgery in both groups (table 2). Frailty will be a mediator in the causal model relationship between prehabilitation and overall quality of recovery.

Outcome measures

Sample size calculation

Group sample sizes of 56 (intervention arm) and 56 (control arm) will achieve 80% power to reject the null hypothesis of equal means when the mean difference is 0.92 (minimal clinically important difference)39 with a SD of 1.72 for both groups taken from our previous study involving cardiac surgical patients18 with the nine-item Quality of Recovery (QoR) questionnaire and with a significance level of 0.05 using a 2-sided two-sample equal-variance t-test. Thus, we wish to detect a medium standardised effect size. To incorporate an expected correlation of 0.5 between prehabilitation and frailty for mediation analysis40 and 10% lost to follow-up, a total sample size of 164 will be needed (82 per group). Our expected correlation of 0.5 is within the results reporting the association between prehabilitation and frailty (r=0.72, 95% CI 0.42 to 0.88).41 We will be using the QoR-15 tool32 as a primary outcome and expect it to perform as well as the nine-item QoR tool we had previously used since equivalence has been established.39 Thus, a clinical meaningful difference of 8.0 with the QoR-15 scale that ranges from 0 to 150 will be expected with the proposed 164 patients.39 This sample size will also have 80% power to detect a probability of 0.626 that prehabilitation has one more DAH₃₀ day (SD 2.2) more than the control group using a Mann-Whitney rank-sum test with a 0.05 two-sided significance level from PASS 14 software (NCSS, Kaysville, UT, USA).

Statistical methods

Missing data will be checked and imputed using the most common category value for categorical variables or median for continuous variables if there is less than 10% missing data. Otherwise, multiple imputation techniques will be used. The Shapiro-Wilk test will be used to check data for normality. Appropriate multivariate tests, with adjustment to type of surgery (covariate used for stratified randomisation), will be used for group comparisons at baseline. All statistical tests will be performed using Stata V.15 with a two-sided significance level set at p<0.05. We will use intention-to-treat analysis for primary analysis, with a per-protocol analysis for sensitivity analysis.

Generalised estimating equation (GEE) models with a Gaussian distribution, identity-link function, unstructured correlation, and robust variance or quantile regression33 will be used to obtain mean differences in primary outcomes between groups after adjusting for covariates (such as type of surgery, age, gender, physical fitness from the 6MWT and comorbidities) based on clinical relevance and adjustments for a directed acyclic graph42 to answer the first study objective. Time-varying confounders may include HADS scores, albumin and haemoglobin concentrations. A GEE model with Poisson distribution and log-link function will be used to obtain a common relative risk of MACCE associated with prehabilitation to help answer the first objective. There is no gold standard for defining frailty, and the relative diagnostic ability of individual scores has not been robustly assessed. While acknowledging this limitation, in order to best answer the second objective, frailty measures will be considered as a mediator variable in the causal model relationship between prehabilitation and recovery measures in the above analyses using techniques outlined by VanderWeele.43 A subgroup analysis between pre-frail (CFS 4) and frail patients (CFS 5 and 6) is also planned to examine if the magnitudes of the association between prehabilitation and quality of recovery differ. To answer another secondary objective, a cost-effectiveness acceptability curve will be drawn to detect differences in the joint cost-effects (DAH₃₀ or QALY) relationships between groups using the net-benefit framework.44 Prehabilitation will be considered as cost-effective if there is a reduction in the overall perioperative treatment cost per gain in DAH₃₀ and/or QALYs.

Patient and public involvement

Patients and the public were not involved in the development of the research question or in the design of the study. Study participants will receive a one-page plain language summary of the results on completion of the study as part of the knowledge translation approach.