Mentored multimodal prehabilitation for aortic aneurysm surgery: a pilot randomised controlled trial

Abstract

Objectives

There is increasing evidence that prehabilitation before surgery may improve patient outcomes and should be established in perioperative care pathways. We aimed to explore the logistics of running a randomised controlled study delivering multimodal prehabilitation (exercise and lifestyle modification) of patients having aortic aneurysm repair. Our primary objective was to assess the feasibility of recruitment, retention, and adherence to study interventions. Secondary objectives were patient outcomes such as function, quality of life and postoperative complications.

Design

A feasibility, assessor-blinded, parallel-group, randomised controlled trial. Following ethical approval, adults scheduled for elective open or endovascular abdominal aortic aneurysm repair were randomised to one of three groups: (1) control, current standard practice with no prehabilitation workshop; (2) non-mentored prehabilitation, a prehabilitation workshop with no further patient contact; and (3) mentored prehabilitation, a prehabilitation workshop with addition of regular mentoring for up to eight weeks after the workshop. Primary feasibility outcomes included: adherence to study interventions (wearing trackers for >80% of the time, attendance to workshop if allocated) and 30% more activity in intervention groups than in control group); screening and recruitment; retention to eight-week follow-up; and maintenance of blinding to the outcome assessor.

Secondary exploratory patient outcomes included complications; mortality; length of stay; quality of life; and functional outcomes.

Results

In total, 60 patients were screened, 45 were enrolled and randomised, and 43 were included in the intention-to-treat analyses. Feasibility of screening and recruitment was achieved, with 45/60 (75%) patients screened being recruited. Participant retention was achieved, with 40/43 (93.0%) of participants attending their eight-week follow-up and 23/43 (53.5%) wearing trackers for >80% of the time, usually due to device failure. Maintenance of blinding was achieved. Participants did not demonstrate a 30% increase in activity levels across the 3 groups. There were no significant differences in secondary patient outcomes.

Conclusions

The primary feasibility targets were partially met in terms of recruitment, retention, wearing of trackers, workshop attendance and completion of follow up procedures. However, the participants did not achieve 30% more activity in the workshop groups compared to the control group. A full randomised controlled trial is feasible with modifications to study design and prehabilitation interventions.

Trial registration

Clinicaltrials.gov ID NCT04169217)

Supplementary Information

The online version contains supplementary material available at 10.1186/s13741-025-00625-6.

Introduction

Prehabilitation has been shown to reduce morbidity and improve postoperative outcomes. Increasing physical activity promotes well-being by reducing stress and anxiety, improving sleep, increasing functional capacity and engagement with surgical teams (Banugo and Amoako 2017; Barberan-Garcia et al. 2018; Gani et al. 2016; Shah et al. 2017; McDowell et al. 2018; Kramer 2020; Levy et al. 2021; Coca-Martinez et al. 2024).

Prehabilitation programmes of varying strategies have been described over the last decade (Valkenet et al. 2011). In particular, programmes associated with major colorectal cancer-related surgery demonstrated encouraging results(, 11Li et al. 2013; undefined), which led to the development of prehabilitation services across healthcare systems, including the NHS (No 2018). The Get It Right First Time (GIRFT) report for vascular surgery published in March 2018 recommended that hospitals undertake prehabilitation programmes (No 2018), but there remains a lack of clarity on methodology or efficacy.

In patients requiring abdominal aortic aneurysm surgery, supervised training programmes have pointed to improve anaerobic, cardiorespiratory fitness levels in research settings(Kothmann et al. 2009; Powell et al. 2014; Tew et al. 2017), and postoperative outcomes, as well as reducing hospital stay, postoperative pain and complications (Barakat et al. 2016). Evidence eluding to safety and slowing of disease progression has been suggested (Tew et al. 2017; Perissiou et al. 2022).

Prehabilitation delivery in vascular surgery pathways, specifically for patients undergoing aortic aneurysm surgery is not as well established. This could be explained by fear of aneurysm rupture, lack of prehabilitation funding in non-cancer services.

Supervised training programmes may not always be practical or desirable for patients who live far from tertiary vascular centres, hence not generalisable nationally. A pragmatic approach to prehabilitation involving mentoring, physiotherapy-led education and exercise workshops, and the use of a smartphone app to promote exercise and lifestyle modifications, may be a more generalisable alternative to supervised training. Ahead of a full randomised controlled trial, we designed a feasibility study. Our primary objective was to assess the feasibility of recruitment, retention, and adherence to study interventions. Secondary objectives were patient outcomes such as function, quality of life and postoperative complications.

Methods

We conducted a single-centre, feasibility, assessor-blinded, parallel-group, randomised controlled trial at a tertiary hospital in London, UK. Our centre performs a large volume of aortic aneurysm repair operations annually, serves a large population of Southeast England, and receives referrals from across the UK. We obtained ethical approval from the NHS Health Research Authority Research Ethics Committee (HRA REC): 19/LO/0988 (NRES Committee, London Central). The study was prospectively registered on clinicaltrials.gov (NCT04169217). Reporting adheres to the CONSORT extension for randomised feasibility studies (Eldridge et al. 2010). The recruitment period was between 20/12/19 and 23/11/22. Funding was provided by Guy’s and St Thomas’ Hospital Charity, Vascular Anaesthesia of Great Britain and Ireland (VASGBI) and The Society and Computing and Technology for Anaesthetists (SCATA).

Patient population

We included adults (aged ≥ 18 years) with an ASA physical status 2 to 4 scheduled for elective open or endovascular (standard and complex) abdominal aortic aneurysm repair. Participants had to be willing to wear an activity tracker for eight weeks and present for follow-up eight weeks after enrolment to undergo assessment. We excluded patients undergoing urgent or emergency surgery, as these patients may not have time to complete study interventions and are more likely to be in-patients, making exercise less practical or safe. We also excluded patients with any severe musculoskeletal disorders preventing participation in exercise (e.g. lower limb amputation). Patients who underwent emergency admission during the eight-week follow-up period were removed from the study at the point of admission.

Primary objectives

Our primary objective was to determine the feasibility of delivering a full trial based on the following criteria:

1. Feasibility of screening and recruitment, with a screening: recruitment ratio of 3:1 (i.e. 25% of screened participants).

2. Subject retention (defined as > 80% of participants attending their eight-week follow-up (virtual or in person), adherence, (wearing trackers for > 80% of the time), and blinding (> 80% of patients having an assessor blinded to their group).

3. Participant adherence to trial intervention (30% more activity in intervention groups than in control group).

We determined that 80% would be an acceptable threshold for wearing trackers. This was selected because we estimated that approximately 10% of the time might be required for battery charging and another 10% for voluntary removal to conduct daily activities. Thus 80% was deemed acceptable and meaningful. Blinding thresholds were also set to 80% as we wanted to ensure that < 3 participants per group had their allocation unblinded, i.e. nine participants in total (20%). Thus, 80% was deemed a meaningful proportion of patients in whom outcome assessors were blinded. Similarly, we assumed that 3 participants per group might not be able to attend their 8-week follow-up (i.e. 9 participants in total). Thus, we expected > 80% to attend as a meaningful threshold. We determined a priori that a 30% change in activity is likely to be a clinically meaningful difference in activity. This threshold of 30% change is used in a variety of settings, including following spinal surgery (Khan et al. 1976; Smith et al. 2020).

Secondary objectives

To determine outcome data that may be used to power a full randomised controlled trial, we also recorded the following: incidence of postoperative complications at 30 days; mortality at 30 days; length of stay; quality of life measured using the EQ-5D-5L(Brazier et al. 2017; Hernandez et al. 2018)at baseline, preoperatively (i.e. following prehabilitation) and six weeks post-surgery; measures of activity and function using standardised measures at baseline, and preoperatively (Duke Activity Status Index (DASI) score, timed up and go (TUAG), gait speed and hip flexion leg muscle strength (Arena et al. 2007; Herman et al. 2011; Mehmet et al. 2020; Thorborg et al. 2010)).

Postoperative complications were defined as having any of the following: cardiovascular complications (myocardial infarction, new onset arrhythmia, prolonged inotropic support for > 12 h postoperatively); respiratory complications (prolonged tracheal intubation > 12 h, re-intubation, postoperative pneumonia); renal complications (need for haemofiltration, increase in creatinine > 20% creatinine from baseline).

Study procedures

Participants were recruited, consented and randomised by a research practitioner into one of three groups in the vascular multi-disciplinary clinic in a 1:1:1 allocation ratio with a block randomisation technique using Sealed Envelope (www.sealedenvelope.com) online software. Participants were allocated to one of three groups: (1) control, (2) non-mentored prehabilitation and (3) mentored prehabilitation. All participants received the standard assessment by the Perioperative care for Older People undergoing Surgery (POPS) team, who provide cognitive, functional and frailty assessments, medical optimisation and health advice.

In group 1 (control), participants received standard care. In group 2 (non-mentored prehabilitation), participants took part in a prehabilitation workshop with no further patient contact. They were given a prehabilitation booklet, including health advice. The workshop was physiotherapist-led and involved education and physical activity training (see below for more information). In group 3 (mentored prehabilitation), participants took part in a prehabilitation workshop with regular ‘mentoring’ up to eight weeks after the workshop, prior to their surgery, and were then given a prehabilitation pack containing health advice. Participants in this group also received a weekly call from a team member to encourage them to increase activity levels as specified in the workshop; a smartphone with a ‘Prehabilitation Activity’ app designed and delivered by Medopad (London, UK), which included motivational notifications, documents and short video clips that we had designed, summarising the importance of prehabilitation and demonstration of exercises; training in how to use the device by the research team; access to expert advice from a designated physiotherapist in normal working hours.

The prehabilitation workshop included lifestyle modifications, such as smoking cessation and dietary advice. Education was provided on exercise; the importance of maintaining physical activity whilst awaiting surgery, the benefits and contraindications to exercise. Participants were educated on the rate of perceived exertion and were advised to work at a moderate activity level. During the prehabilitation workshop, groups participated in a strength and balance exercise regime. This consisted of marching on the spot interspersed with lower limb strengthening exercises; hip abduction, squats, hip extension, calf raises. Participants were advised to complete two sessions of strength and balance exercises per week, as demonstrated in the group, and three sessions of cardiovascular exercise per week, aiming for 30 min. Participants were encouraged to move regularly and limit their time spent sedentary. The prehabilitation groups were also counselled on what happens during their inpatient stay and set expectations regarding early mobilisation postoperatively. The third group, in addition to the workshops, received weekly mentoring by the research team, direct access to a physiotherapist, and encouraging push notification messages on the phone. This continuous support was designed to make the participants feel cared for and supported throughout their prehabilitation journey. Safety measures surrounding falls were added due to the remote element of the workshop. A risk assessment for each patient was carried out prior to engaging in the physical activity component of the workshop.

Adjustments to study protocol due to COVID-19 pandemic changes

Due to the outbreak of the COVID-19 pandemic, study interventions involving prehabilitation workshops were converted to a remote online delivery using a virtual platform. The planned eight-week EQ-5D-5 L follow-up questionnaires were carried out via virtual means instead of face-to-face. Eight-week functional follow-up assessments were non-mandatory. All changes underwent ethical review and approval before implementation.

Activity trackers

All patients had their activity tracked using an Actiwatch Spectrum Plus (PHILIPS Healthcare, Eindhoven, Netherlands) activity tracker. This is a lightweight, waterproof, wrist-worn device. Participants wore the activity tracker for eight weeks. After completion of the eight-week intervention period, participants returned the tracker in person during their admission for surgery. All data were stored on the trackers and downloaded anonymously onto Respironics Actiware 6 software (PHILIPS Healthcare) and stored onto secure research computers for analysis.

The Actiwatch Spectrum Plus device measures activity counts using an accelerometer. Activity counts, mobile time, time off wrist and total mobile bouts were measured and compared between the study groups. The device records motion and light and has a sensor to detect when being worn. A count is measured as the frequency response range (e.g. 0.35–7.50 Hz) and integrates data as a peak acceleration detected over each epoch (Chen and Bassett 2005). An epoch length is the period of time the Actiwatch will accumulate activity counts. The epoch was set at 1 min for all patients; at the end of each epoch, physical activity (counts) were calculated, and the process was repeated until the data collection was completed (Ayabe et al. 2013).

For missing activity tracker data, it was assumed that subjects were either not wearing their trackers or had a technical failure in capturing data. These missing data were conservatively imputed as zero values (Yi et al. 2023).

Follow-up

Patients were followed at two time points in the trial: before surgery after the eight-week intervention period (follow-up 1) and six weeks postoperatively (follow-up 2).

Follow-up 1: eight weeks after recruitment, before surgery

Participants were contacted to complete an EQ-5D-5 L quality of life questionnaire. Pre-operative physical functional assessments (secondary outcomes) were carried out if practical due to the COVID-19 pandemic, but due to access restrictions during the pandemic, participants were not specifically brought into the hospital for functional assessments only.

Follow-up 2: six weeks postoperatively

Participants were contacted by telephone to complete an EQ-5D-5 L quality of life questionnaire.

Electronic patient records and electronic discharge letters were used to collect data on postoperative complications, 30-day mortality and length of hospital stay.

Blinding

Outcome assessors were blinded to the intervention as the intervention was performed remote to data collection; thus outcome assessors were unable to determine what group participants in. The statistician had no access to source data, and groups were anonymised on data collection forms. No team members other that the chief investigator were able to access the randomisation allocations.

Statistical analysis

A convenience sample of 15 patients per group was planned, with a total of 45 patients recruited. This was deemed acceptable to yield sufficient data to help future power analyses for a larger trial. We took into account practical considerations such as participant flow, funding constraints, and the number of participants needed to reasonably evaluate feasibility goals(Yi et al. 2023), with other feasibility studies published with similar recruitment numbers per group (Coca-Martinez et al. 2024; Teresi et al. 2022).

Data were collected using a standardised case record form and compiled into a registered electronic database Castor, (Castor, Amsterdam, The Netherlands). Primary data analyses used intention-to-treat principles, with per-protocol analyses reported in the online Supporting Appendix. Completion of intervention is defined as having completed all steps, including wearing the activity tracker for more than 80% of the defined period, completed all of the follow-up steps and having had their surgery. The excluded patients in the per-protocol analysis were those who lacked activity tracker data for more than 20% of the defined period and or those who did not undergo surgery.

Data analyses were performed using R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). Normality testing of continuous variables was conducted using the Shapiro-Wilk test. Normally distributed data were reported as mean (standard deviation (SD)) and assessed for differences in means between groups with a one-way analysis of variance test (ANOVA), whilst non-normally distributed data were analysed using the Kruskal-Wallis Rank Sum Test. Categorical data was assessed using chi-squared tests. A p-value of < 0.05 was considered significant.

Results

A total of 60 patients were screened for enrolment, of which 45 were enrolled and randomised into the three study arms and 43 patients were included in intention-to-treat analyses after two patients withdrew consent following randomisation (Fig. 1). A total of 23 patients completed all parts of the study protocol and were included in secondary per-protocol analyses. Table 1 summarises the characteristics of participants included in the intention-to-treat analyses (see Appendix 1).

Fig. 1.

CONSORT flow diagram. Sixty patients were screened for the study and 45 consented and randomised. Two patients were excluded from intention-to-treat analysis due to withdrawal of consent

Table 1.

Participant characteristics demographics in the intention to treat cohort. Values are mean (SD), number (proportion) or median [IQR]

	Overall (n = 43)	Control (n = 14)	Non-mentored Prehabilitation (n = 14)	Mentored Prehabilitation (n = 15)
Age (years)	68.47 (6.48)	70.16 (6.21)	66.79 (6.72)	68.46 (6.51)
Ethnicity
Black	1 (2.3%)	1 (7.1%)	0 (0.0%)	0 (0.0%)
Other	1 (2.3%)	1 (7.1%)	0 (0.0%)	0 (0.0%)
White: British	37 (86.0%)	11 (78.6%)	12 (85.7%)	14 (93.3%)
White: Irish	1 (2.3%)	0 (0.0%)	1 (7.1%)	0 (0.0%)
White: Other	3 (7.0%)	1 (7.1%)	1 (7.1%)	1 (6.7%)
Sex (male)	36 (83.7%)	11 (78.6%)	11 (78.6%)	14 (93.3%)
Weight (kg)	85.17 (17.85)	77.01 (13.13)	89.66 (19.23)	88.60 (18.80)
Height (m)	1.73 (0.08)	1.69 (0.07)	1.73 (0.07)	1.76 (0.08)
Body mass index (kg/m2)	28.48 (4.97)	27.02 (3.78)	29.80 (5.49)	28.63 (5.38)
Aneurysm size (cm)	6.00 [5.80, 6.40]	6.00 [5.65, 6.27]	6.15 [6.00, 6.47]	6.00 [5.80, 6.30]
Aneurysm site
Abdominal: Infrarenal	17 (39.5%)	8 (57.1%)	5 (35.7%)	4 (26.7%)
Abdominal: Juxtarenal	19 (44.2%)	3 (21.4%)	6 (42.9%)	10 (66.7%)
Abdominal: Suprarenal	3 (7.0%)	0 (0.0%)	2 (14.3%)	1 (6.7%)
Thoracic	1 (2.3%)	1 (7.1%)	0 (0.0%)	0 (0.0%)
Thoracoabdominal	3 (7.0%)	2 (14.3%)	1 (7.1%)	0 (0.0%)
Comorbidities
Hypertension	26 (60.5%)	9 (64.3%)	10 (71.4%)	7 (46.7%)
Peripheral vascular disease	4 (9.3%)	0 (0.0%)	2 (14.3%)	2 (13.3%)
Cardiovascular disease	14 (33.3%)	3 (21.4%)	5 (35.7%)	6 (42.9%)
Respiratory disease	22 (51.2%)	7 (50.0%)	5 (35.7%)	10 (66.7%)
Diabetes	10 (23.3%)	2 (14.3%)	3 (21.4%)	5 (33.3%)
Renal disease	5 (11.6%)	1 (7.1%)	0 (0.0%)	4 (26.7%)
Smoking	22 (51.2%)	7 (50.0%)	5 (35.7%)	10 (66.7%)

Primary outcomes

Feasibility of screening and recruitment

Between December 20, 2019, and November 23, 2022 60 patients were screened for eligibility to participate in the study. Of these, 45 patients (75%) were recruited and randomised, and 43 patients (71.7%) were included in the intention-to-treat analysis. There were low numbers of missing variables for baseline patient characteristics.

Retention and adherence to eight-week follow-up and blinding

In total, 40/43 (93.0%) patients attended the eight-week follow-up. All patients who recorded data every week attended this follow-up appointment. Watches were sent back to the research team and the activity tracker data from these were analysed for each participant. Missing data were present for 92/344 participant-weeks (26.7%). On weeks when activity tracker data were missing, it was assumed that patients either did not wear the tracker for the entire week or that the tracker failed. Adherence to wearing the activity trackers was incomplete with 23/43 (53.5%) wearing trackers for more than 80% of the time.

All independent assessors successfully maintained blinding throughout the study.

Activity counts

Activity counts were summed for each patient over the eight-week period, and on weeks where activity tracker data were unavailable, it was assumed that the activity count that week was zero according to intention-to-treat. The median (IQR) activity count was 5,772,748 (3,214,105–8,339,418) across all patients. There was no statistically significant difference in activity counts between groups (Table 2; Fig. 2). On a per-protocol analysis, the median (IQR) activity count was 7,287,085 (5,767,559–9,150,332) across all patients (Appendix 1, Table S1 and Fig. S1).

Table 2.

Outcomes in the intention-to-treat cohort. Values are median [IQR] or mean (SD). p-values represent the probability under a null hypothesis of no difference in distributions between the 3 groups in a one-way ANOVA (for normally-distributed data) or a Kruskal-Wallis rank sum test (for non-normally distributed data)

	Overall (n = 43)	Control (n = 14)	Non-mentored prehabilitation (n = 14)	Mentored prehabilitation (n = 15)	p value
Activity counts
Total activity count	5,772,748 [3,214,105 to 8,339,418]	5,581,653 [3,365,584 to 7,289,536]	7,057,828 [5,179,980 to 9,392,906]	5,371,694 [1,461,025 to 7,493,204]	0.222
EQ-5D-5 L Visual Analogue Scale
Baseline	80.00 [65.00 to 95.00]	82.50 [75.00 to 93.75]	92.50 [68.75 to 95.00]	70.00 [52.50 to 82.50]	0.141
Pre-operative	80.00 [75.00 to 95.00]	90.00 [75.00 to 99.00]	80.00 [80.00 to 90.00]	75.00 [65.00 to 85.00]	0.118
Postoperative	80.00 [60.00 to 95.00]	77.50 [52.50 to 93.75]	87.50 [76.25 to 93.75]	75.00 [60.00 to 85.00]	0.294
DASI Scores
Baseline	48.45 [38.95 to 58.20]	50.20 [45.08 to 58.20]	54.45 [38.01 to 58.20]	42.70 [30.83 to 50.70]	0.248
Pre-operative	41.20 [30.01 to 50.70]	42.70 [31.45 to 50.70]	47.20 [34.70 to 58.20]	36.70 [26.70 to 42.70]	0.242
TUAG Scores (s)
Baseline	10.06 (2.47)	10.89 (2.98)	9.29 (1.60)	10.01 (2.54)	0.235
Pre-operative	9.60 (2.00)	10.23 (2.60)	9.00 (1.53)	9.61 (1.69)	0.318
Gait speed (m/s)
Baseline	1.09 (0.26)	1.06 (0.31)	1.10 (0.20)	1.09 (0.27)	0.932
Pre-operative	1.09 (0.21)	1.06 (0.25)	1.12 (0.21)	1.10 (0.20)	0.784
Leg muscle strength
Baseline Force (kg)	19.05 [15.60 to 22.57]	18.10 [13.72 to 22.54]	22.26 [19.08 to 26.05]	17.60 [15.60 to 20.66]	0.076
Pre-operative Force (kg)	20.71 [18.10 to 23.20]	19.30 [17.60 to 21.98]	20.71 [18.20 to 23.87]	21.62 [18.75 to 22.65]	0.699
Baseline Peak Force (sec)	4.11 [3.41 to 4.57]	4.27 [3.84 to 4.54]	4.12 [3.71 to 4.62]	3.83 [3.00 to 4.39]	0.359
Pre-operative Peak Force (sec)	3.79 [3.24 to 4.22]	3.72 [3.51 to 4.29]	3.84 [2.84 to 4.02]	3.97 [3.66 to 4.54]	0.397

Fig. 2.

Boxplots of activity counts in the intention-to-treat cohort. The distribution of activity counts in each group is shown. Each dot represents the total recorded activity count of a patient during the study period. The hinges of the box represent the interquartile range and the dark line in each box represents the median for that group. The whiskers represent 1.5 times the interquartile range. Black dots represent outliers with recorded values beyond 1.5 times the interquartile range

Secondary outcomes

The incidence of postoperative complications in the intention-to-treat cohort was 5/43 (11.1%) and 2/23 (8.7%) in the per-protocol cohort (Appendix 1, Table S3). The median (IQR) postoperative length of stay in the intention-to-treat and per-protocol cohorts were 7 (4–10) days and 7 (4–11) days, respectively.

There was no statistically significant difference in EQ-5D-5 L Visual Analogue Scale between groups detected in either intention-to-treat (Table 2; Fig. 3) or per-protocol cohorts (Appendix 1, Table S4 and Fig. S2). There were no differences between groups in terms of EQ-5D-5 L subdomain outcomes, apart from mobility, where a larger proportion of participants reported a deterioration in their mobility in the mentored prehabilitation group at the pre-operative visit over baseline, compared with other groups (Appendix 1, Tables S4 and S5). There were no differences in DASI, time up and go, gait speed and leg strength between the groups (Table 2 and Appendix 1, Table S1).

Fig. 3.

Boxplots of EQ-5D-5 L visual analogue scale in the intention-to-treat cohort. The scores are plotted for each group at three time-points in the study: at baseline, at eight weeks after recruitment prior to surgery (pre-operative), and six weeks postoperatively (postoperative). The hinges of the box represent the interquartile range and the dark line in each box represents the median for that group. The whiskers represent 1.5 times the interquartile range. Black dots represent outliers with recorded values beyond 1.5 times the interquartile range

No deaths occurred within 30 days of surgery. Ten serious adverse events were reported, one of which resulted in a protocol deviation requiring the participant to be excluded from the study, and none were deemed to be related to the study interventions (Appendix 3).

Discussion

Main findings

Delivering and integrating multimodal prehabilitation into clinical services is challenging (Boyle et al. 2023). This study was an initial step in designing a full randomised controlled trial and planning an effective and practical prehabilitation programme for patients with planned aortic aneurysm surgery. This feasibility study aimed to assess adherence to the study protocol and delivery of our prehabilitation intervention, and demonstrated acceptable screening, recruitment and retention. However, there was no significant differences between the level of activity between groups suggesting that our supervised one-off workshop might be insufficient to increase pre-operative patient activity.

More than half of the study participants adhered fully to protocol, specifically wearing of the Actiwatch and most performed follow-up procedures. Technical failures of the Actiwatches were the primary cause of data loss rather than the unwillingness of patients to wear them. Anecdotally, patients did not always have the capacity or ability to log onto the online physiotherapist-led workshops. There was a low participant dropout rate, demonstrating participants’ willingness to enrol and complete research such as this. Primary and secondary outcome measures collected in the study were straightforward and easy to capture. In particular, the data collected at follow-up 1 (eight weeks following randomisation pre-operatively), and follow-up 2 (six weeks post-operatively) exhibited low amounts of missing data and good adherence to protocol. There was no significant difference in amount of physical activity between the study groups.

This was a feasibility study and not powered to detect differences between groups, although no clinically or statistically significant differences were found between the groups in the incidence of composite of postoperative cardiac, respiratory and renal complications at 30 days; mortality at 30 days; length of hospital stay; quality of life measured using the EQ-5D-5 L, DASI, gait speed, TUAG. The fact that QoL domains demonstrated a reduction rather than improvement may serve to highlight that patients may in fact deteriorate while waiting for surgery, or it may be that patients may experience questionnaire fatigue, and answer out of frustration. Importantly, the sample size was not large enough to draw any meaningful conclusions.

Although the existing body of research demonstrates proven benefits on physiological and other patient centred outcomes, the findings of this study does not support this due to it being an under powered feasibility study. The results of this study however, do support the some of the primary outcomes related to recruiting and adhering to study interventions. There was no significant differences between the level of activity between groups suggesting that our supervised one-off workshop is not a powerful enough intervention to encourage increased activity levels.

Context to the intervention

Patients are often erroneously told to reduce physical activity and exercise when diagnosed with an aneurysm. Recent studies have shown exercising with an aneurysm is safe and does not increase risk of rupture or sac size(Tew et al. 2017; Barakat et al. 2016). Our study design allowed for a one-off workshop, which relied on a large educational component to exercise prescription with a heavy focus on decreasing sedentary behaviour and increasing general physical activity. The aim was to debunk the myth of exercising with an aneurysm being unsafe. However, this relies on the patients having a good understanding of exercise perception and rate of perceived exertion. Due to the nature of a one-off workshop the exercises had to fit a wide range of capabilities. The exercises used within the study were largely based on the Otago exercise programme which has been shown to improve strength and balance in the elderly (Yi et al. 2023) alongside the World Health Organization recommendation for adults over 65 years. For some participants, these exercises may not have been challenging enough to make meaningful changes in strength and cardiovascular fitness. Of those who received regular mentoring, few required physiotherapy advice regarding the exercises prescribed, and most gave positive feedback regarding adherence.

Challenges and solutions

This preliminary study underscored several technological hurdles that necessitate attention in any subsequent randomised controlled trial and in the development of a comprehensive prehabilitation programme. It is imperative that technologies, such as activity trackers, exhibit durability and dependability. Another adjustment to a future study would be to objectively measure exercise intensity during the exercise sessions and not activity counts as per this pilot study. This is important to establish if the exercise component of any programme is adequate enough to achieve the desired effect. Providing ample support is essential to ensure that patients derive optimal benefits from technological applications like prehabilitation apps and videoconferencing software. A significant challenge lies in addressing the disinterest in technology and the substantial level of digital illiteracy within the elderly vascular populations (Lareyre et al. 2022). The optimal remedy to this is unclear; older age and socioeconomic status are contributory factors that may prove difficult to overcome (Chowdhury et al. 2023). A hybrid prehabilitation approach could potentially offer a resolution, delivering home-based or in-person interventions according to patient needs.

We faced challenges in identifying patients and embedding a prehabilitation programme within a variable surgical timeline. Often a patient would be recruitable, but changes to their surgical approach while awaiting a multidisciplinary team (MDT) decision would extend the time to operation. The median (IQR) time to surgery was 113 (61.25–188.25) days. Overall, 34 patients (76%) exceeded the anticipated 8-week timeline to surgery.

The main change would be related to inclusion criteria. It may be easier to recruit participants having one type of aortic aneurysm repair, e.g. standard open surgical repair of abdominal aortic aneurysm. The lead times for more complex endovascular aneurysm repair can be prolonged while awaiting MDT decisions or bespoke graft design. In contrast, open surgical repairs are usually done within more clearly delineated timeline and should be treated within eight weeks (Davis et al. 2013). This would still fit in within the ideal prehabilitation timeline of four to eight weeks (Shakya and Poudel 2022).

Elective open abdominal aortic aneurysm repairs are generally performed in younger and fitter patients, which would make the use of technology and virtual platforms more appealing. Generally, there is a shorter surgical lead time, which would make data collection and follow-up procedures more suitable. Open abdominal aortic aneurysm surgery is a larger physiological insult to the patient in the perioperative period. Recently published data suggest that open repairs may have better durability and benefits in the longer term(Vallabhaneni et al. 2024), highlighting an opportunity for prehabilitation to be established in open aneurysm surgery pathways. This may be explained by the fact that patients having abdominal incisions are more likely to have postoperative surgical complications, making the distinction between the two groups easier to detect. The disadvantage of selecting only open abdominal aortic aneurysm repairs would be longer recruitment times and the exclusion of older patients who tend to be offered the endovascular surgical approach. In our unit, endovascular repair is the predominant choice for the > 75 years age group. This would mean that the study might not be generalisable to the whole aortic aneurysm population and introduce a selection bias.

The data collection and outcomes selected for this feasibility study were straightforward and achievable. However, around a quarter of activity tracker data were missing, suggesting that wearable device technology may have acceptability and reliability issues which need to be overcome for further studies.

COVID- 19 implications

The impact of COVID-19 had significant and diverse repercussions on the trial. Non-urgent operations and appointments were cancelled, leading to a halt in trial recruitment for several months. Diminished access to healthcare, reduced physical activity, increased levels of depression, and increased reliance on alcohol and smoking globally had an adverse effect on the physical and mental well-being of the population and possibly our participants (Hunter et al. 2023). Additionally, in order to facilitate physiotherapist-led workshops, we had to incorporate an unplanned virtual element to the trial in the face of the pandemic, necessitating modification of the protocol.

The pandemic meant that we had to change our approach from face-to-face physiotherapy-led workshops to a virtual approach, which was useful for patients who had the means and lived far away. Changing to a virtual platform could be an advantage, however, it may disadvantage those with restrictions to using the technology. In our pilot, 4/45 participants did not have or want to engage with the technology.

Slow recruitment was multifactorial. In particular, during the trial’s delivery, significant changes were made to surgical care pathways locally and regionally for obvious reasons. As a tertiary centre receiving referrals from across the region, this hampered prompt identification of patients and added organisational challenges beyond just clinical research. Importantly, availability of research staff and our ability to re-open trials was limited for extended periods of time.

We noted that there was initially low uptake into the study to due to lack of access to technology, and therefore during the course of the trial we secured funding for internet-equipped tablets and wrote explanatory material for participants to be able to engage further. The tablets were acquired February 2022. From that point nine participants were randomised into workshop groups and four participants out of the nine needed to borrow the tablets to attend virtual workshops. This improved engagement and mitigated the risk of health inequity, and should be considered in future studies.

See Table 3 for future study interventions.

Table 3.

Table describing the study interventions for a future randomised controlled trial

Future Study Interventions
Eligibility	Elective infrarenal open abdominal aortic aneurysm meeting threshold for surgery > 60 mm Participants should undergo a minimum 4 weeks of prehabilitation
Randomisation	No prehabilitation (standard care) v supervised sessions 3–4 times a week. The standard care group is standard pre-operative assessment which includes exercise and lifestyle advice.
Prehabilitation interventions	Supervised exercise training Face to face and a virtual platform for the exercise sessions. Consider splitting the intervention group into moderate intensity and high intensity groups to assess outcomes in each level of intensity.
Activity trackers	Use of reliable wearable trackers with long battery life or easy to charge.
Monitoring of effect	Measure objective exercise intensity during sessions to ensure participants reach high intensity for a significant portion of the session. Use the Karvonen formula and heart rate monitoring.

Limitations

There were several limitations to this study. First, there was no health economics or implementation science analysis, which would be beneficial for a future randomised trial. We did not objectively and explicitly seek usability and acceptability data on the fitness tracker devices. However, we recognise that technology has evolved in the recent past, and therefore novel devices might be more usable and acceptable in future clinical trials. We selected an eight-week window of activity, primarily because this is the usual waiting time for complex aortic graft design and manufacture. It is unclear if extending this time would provide additional benefit to physical activity and prehabilitation outcomes, but this will warrant future investigation. The demographic recruited into the study were mainly men from a White British background, and therefore the results should be interpreted as mostly applicable to this cohort of patients.

We did not stratify participants to surgical technique for this pilot as this would have slowed down the recruitment process further and meant that we would need larger numbers. This would certainly have implications of post-surgical outcomes and would have to be accounted for in a full randomised controlled trial. Finally, we chose a 30% threshold as a marker of increased activity. Although this threshold is applicable to certain populations, it has not been validated in the cohort of interest in this study.

Conclusions

This feasibility study has demonstrated the potential to deliver a full randomised controlled trial, with modifications to the study processes to address limitations. It has provided an opportunity to refine study procedures to improve the possibility of determining the impact of prehabilitation on outcomes in the vascular population having aortic aneurysm repair.

Authors’ contributions

HB: Study design, recruitment, consent and write up of manuscript.DW: Statistical analysis data presentation and contribution to the write up.GC: Recruitment, consent, randomisation and data entry.PB: Study design, design of prehab material and write up of the manuscript.HW: Conduct of prehab intervention and follow up data collection.LZ: Study design and securing funding.MB: Conduct of prehab intervention, design of participant information leaflets and guidelines for patients and study interventions.KEB: Study design and write up of manuscript.All authors read and approved the final manuscript.

Funding

Funding was secured to run the study from the following sources: uy’s and St. Thomas’ charity, UK Society for Computing and Technology in Anaesthesia (SCATA) and Vascular Anaesthesia Society of Great Britain and Ireland (VASBGI).

Data availability

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Declarations

Ethics approval and consent to participate

Ethical approval was sought from NHS Health Research Authority Research Ethics Committee (HRA REC): 19/LO/0988 (NRES Committee, London Central).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.