Neuromuscular electrical stimulation as an adjunct to standard care in improving walking distances in intermittent claudication patients: the NESIC RCT

Article history

The research reported in this issue of the journal was funded by the EME programme as project number 15/180/68. The contractual start date was in November 2017. The final report began editorial review in January 2022 and was accepted for publication in August 2022. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The EME editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

Copyright © 2023 Burgess et al. This work was produced by Burgess et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2023 Burgess et al.

Background: intermittent claudication

Peripheral arterial disease (PAD) is the chronic obstruction of the arteries supplying the lower limbs caused by atherosclerosis. The incidence of PAD increases with age, and in the United Kingdom (UK) around one in five people over the age of 60 have some form of PAD. 1 Risk factors include smoking, hypercholesterolaemia, hypertension and diabetes. These individuals are more likely to suffer comorbid conditions such as heart attacks and stroke. 2

Intermittent claudication (IC) is the most common manifestation of symptomatic PAD, presenting as pain or weakness with walking that is relieved with rest. This is functionally debilitating and results in a poor quality of life (QoL). 3 IC symptoms remain stable for the majority of patients but around 5–10% may develop critical limb ischaemia (CLI). CLI is characterised by a severe obstruction in the circulation of the lower extremities, ischaemic pain and tissue loss (gangrene/ulceration). In some cases, this may eventually lead to limb amputation, with associated changes to QoL. Those patients with diabetes are at a higher risk. In the UK, PAD is the single largest cause of limb amputation. 1

Treatment options for intermittent claudication

The National Institute for Health and Care Excellence (NICE) guidelines recommend that patients with IC should be offered support and treatment regarding the secondary prevention of cardiovascular disease. This includes exercise advice (EA), lipid modification and statin treatment, antiplatelet therapy as well as the prevention, diagnosis and management of high blood pressure and diabetes, known as best medical therapy (BMT). They should also be offered a supervised exercise therapy (SET) programme as a first-line treatment option. 4 SET classes usually involve a circuit of lower-limb exercises under the supervision of a health-care professional, for a minimum of 30 minutes per week usually over 3 months duration. Only if BMT and SET have not led to a satisfactory improvement in IC symptoms is surgical intervention offered for suitable patients (angioplasty, primary stent placement or bypass). 4

There are a number of vasoactive drugs licensed to treat the symptoms of IC specifically when conservative treatment has been ineffective, with NICE recommending naftidrofuryl oxalate as the preferred treatment4 as it is the most cost-effective and efficacious (up to 60% improvement). 5

Summary of current research

There is a strong evidence base supporting the initial management of IC as per NICE guidance,4 including BMT and SET to increase the pain-free walking distance. 6

A Cochrane systematic review of the impact of SET on walking time or distance was carried out by Bendermacher et al. in 2006,7 repeated in 2013 by Fokkenrood et al.,6 and updated again in 2018 by Hageman et al. 8 The latter review included parallel-group randomised controlled trial (RCT) data comparing SET to home-based exercise therapy and walking advice in patients with IC. Twenty-one studies with a total of 1400 participants were randomised and followed up between 6 weeks and 2 years, with the primary outcome measure being maximal walking distance or time (MWD/T). There was a significant improvement in MWD/T compared with home-based exercise therapy and walking advice, with overall standardised mean differences at 3 months of 0.80 [95% confidence interval (CI) 0.53 to 1.07; p < 0.00001; high-quality evidence]. This translates to an improvement in MWD of 210 m in the SET group. 8

Despite its beneficial effects, SET is underutilised in the UK. Recommended care for the first-line management of claudication is significantly below standard largely due to lack of National Health Service (NHS) funding. In 2017, 89 Vascular Society of Great Britain and Ireland members completed a survey, representing 59 (57%) of the 97 vascular units registering data on the National Vascular Register. Of the respondents, 37 (41.6%) members reported that they had access to SET, which equates to only 22 (38.5%) of the vascular units having access to a supervised exercise programme for IC patients. 9 A 2021 audit10 showed that only 36% of UK vascular units have access to SET for PAD patients, and only four are fully compliant with current NICE guidelines. With increasing constraints on NHS budgets, poor access to SET is unlikely to improve. Where available, the authors noted poor uptake and adherence, with reasons including lack of transportation to SET centres, personal travel expenditure, inflexibility of classes and absence from work being cited.

To attend SET for 2 hours per week for 3 months duration costs approximately £288 per patient, equating to approximately £1608 per quality-adjusted life-year (QALY) gained. 3 This includes the time of a physiotherapist or allied health-care provider supervising within the physiotherapy gymnasium with equipment including a treadmill, steps and walking cones. This is for a finite treatment period as dictated by the SET and does not include the patient’s own costs. 3

Clinical practice is variable between clinicians for prescribing vasoactive medications. 11 Additionally, their efficacy in clinical trials has been variable,12,13 there are associated side effects such as diarrhoea and vomiting14 and they are contraindicated in certain conditions such as hyperoxaluria or recurrent calcium-containing stones. 15 A systematic review by Momsen et al. 13 concluded that there are a number of drugs that improve MWD but with limited benefits.

Without a demonstrable benefit of non-invasive strategies for the management of IC, there is an increased likelihood of invasive treatment options. These procedures are expensive, for example bypass surgery costs approximately £8857.00, procedure cost. 16 Additionally, patients undergoing surgery have more complications and so may be more of a clinical and economic burden on the NHS. 17 A cost-effectiveness study conducted by Djerf et al. 18 concluded that the costs of revascularisation in conjunction with BMT in IC patients were approximately four times higher than for those receiving BMT alone. The incremental cost-effectiveness ratio (ICER) of revascularisation exceeded that of the NICE guidelines. 18

The true standard of care, therefore, for the majority of patients with IC in the UK and Ireland is BMT only. Therefore, adjuncts to these therapies must be explored.

Neuromuscular electrical stimulation

Emerging technologies include neuromuscular electrical stimulation (NMES) devices, which may be beneficial in some people suffering with IC by improving the distance walked before symptomatic limitation and improve QoL. 19 While evidence is limited, a systematic review of five trials conducted by Williams et al. 20 investigating different NMES devices used as a treatment option for IC patients demonstrated an improvement in MWD by up to 150% at 4 weeks of intervention.

Moreover, a proof-of-concept pilot study of 20 participants with IC showed a significant improvement in absolute walking distance (AWD) of 85 m (102.3 m vs. 187.2 m, p < 0.01) and initial claudication distance (ICD) of 38 m (50.5 m vs. 88.2 m, p < 0.01) in a 6-week period. Using a REVITIVE IX (Actegy Health Ltd, Bracknell, UK) device, all patients underwent 30 minutes of NMES daily at their own convenience, in the comfort of their homes. Repeated measures were then taken at the 6-week follow-up appointment. In addition to this functional improvement, there were significant improvements in both validated generic EQ-5D-5L scores (0.5427 vs. 0.6443, p < 0.005) and disease-specific intermittent claudication questionnaire (ICQ) (44.3 vs. 35.21, p < 0.002) QoL questionnaire scores at 6 weeks. 19 Compliance to the device during the 6-week intervention period was 98.5% as assessed by patient-recorded diaries. A subsequent RCT compared SET (group A) versus SET plus NMES (group B). The AWD and ICD both significantly increased over the 6-week treatment period in both groups, with the change in ICD in group B being significantly greater than that in group A (40.4 vs. 7.5 m, respectively; p = 0.012). 21

Technological advances have allowed portable, inexpensive and safe electrical-stimulation units to be developed which can be used in the patient’s own home. 22 These devices deliver therapeutic levels of intensity to cause contraction of the calf muscles in similar ways to intermittent limb compression and may have similar beneficial effects. 20

Mechanistic evaluation of the device

This study also aimed to evaluate the potential underlying mechanism by which NMES may improve walking distances in patients with IC. A number of devices that perform compression have been developed, the most common of these being intermittent pneumatic compression devices. Studies evaluating such devices have shown functional and symptomatic benefits in patients with IC. 20 These work by applying high-level pneumatic compression to the foot and/or calf, reducing the venous leg pressure and consequently increasing flow rate in the popliteal artery and stimulating the release of vasodilators. 23,24 It is hypothesised that these physiological responses are responsible for improving claudication symptoms. This mechanism of action may be similar for NMES devices. The RCT by Babber et al. 21 found significant increases in volume flow (VF) and time average mean velocity (TAMV) when the device was switched on at baseline and at week 6, although this was not maintained after device cessation.

Further haemodynamic assessment is therefore required in this study to help better understand and assist in developing future technology to optimise the use of this mechanism for patient benefit.

Rationale for the NESIC study

Supervised exercise, BMT, medications and radiological and surgical intervention are all effective therapies20 for treating IC. However, mortality rates related to PAD are rising,20 with a death rate of 20% within 5 years following diagnosis,25 and there are limitations to their use.

Current NICE guidelines4 for the initial management of IC are impractical. Although evidence-based, there is a significant underutilisation of SET, an evidence-based treatment modality that can significantly improve functional ability. 6 The underutilisation is driven by a chronic lack of NHS funding to support staff and set up resources for an exercise programme as well as due to compliance, as patients often need to travel long distances at their own expense on a regular basis in order to benefit from attending an exercise class. Patients who are in employment often decline an invitation for SET, do not attend or require significant time off work to attend. The more realistic picture of current initial management of IC is BMT only.

Invasive procedures such as bypass or angioplasty (using a balloon to widen a narrowed artery) to restore blood flow carry risks of operative complications and often patients are unsuitable for such interventions. 20 These procedures are also expensive and if these measures fail, a major amputation is the usual fate, with associated changes to QoL. 26

An effective, non-invasive modality that will promote compliance with treatment or act as a valid alternative for the majority unable to access SET is therefore required. The per-unit price of a commercially available NMES device is approximately £250 and therefore cheaper than the per-person cost of attending SET, which is also limited to a treatment duration of 3–6 months. 6 The NESIC trial aimed to determine:

Is there an adjuvant benefit of NMES to locally available therapy, including SET or BMT only? Is NMES cost-effective in this role compared to SET? Is there potential for NMES use as first-line management of patients with IC?

Primary objectives

The primary objective was to compare the mean difference in AWD at 3 months in patients with IC receiving a NMES device and local standard care (intervention), compared with local standard care alone (control).

Secondary objectives

Other objectives included:

• change in ICD

• compliance with NMES during the 3-month treatment period

• compliance with the localised SET programme at SET centres

• to understand the underlying mechanisms for change in clinical and subjective outcomes in the form of lower-limb gross and superficial haemodynamic assessment

• QoL – change in European Quality of Life 5-Dimension 5-Level (EQ-5D-5L^®) (EuroQol Group, Rotterdam, The Netherlands) and Short-Form Health Survey-36 (SF-36^®) (RAND Health Care, Santa Monica, CA, USA) (validated generic QoL tools) and the ICQ over 12 months from baseline

• to assess the cost-effectiveness of the NMES device compared to SET.

Trial design

A multicentre, pragmatic, randomised clinical trial to compare the mean difference in AWD at 3 months from baseline in patients with IC. Participants were randomised 1 : 1 to either:

1. local standard care (control)

2. NMES device and local standard care (intervention).

Changes to the trial design

The NESIC trial aimed to recruit 96 patients in each arm (192 in total): SET arm (96 patients) and non-SET arm (96 patients). The SET recruitment target (96) was met in June 2019 but continued in order to replace participants who had been excluded post randomisation (108 SET patients were successfully randomised in total).

Recruitment into the non-SET arm continued and was extended until 31 March 2020. This extension of recruitment was approved by the National Institute for Health and Care Research (NIHR) Efficacy and Mechanism Evaluation (EME), and they advised to wait to submit the contractual agreement until after recruitment had completed so actual amounts requested could be confirmed However, due to the COVID-19 crisis, recruitment was formally paused early on 20 March 2020. At this point 92 non-SET patients had been randomised and only four more patients were required to meet the recruitment target. Following advice from the Independent Data Monitoring Committee (IDMC) and the Trial Steering Committee (TSC), recruitment did not restart and was formally closed as it was not deemed worthwhile to keep recruitment open in the existing COVID-19 climate for four more participants and it was likely that sufficient power had already been achieved. This was discussed with the EME Programme Director, who supported the decision to not reopen recruitment. A total of 200 patients were successfully randomised into the study.

In November 2020 the EME programme team approved a contract variation, awarding an 8-month extension to enable sites to continue to exclusively recruit non-SET participants and to ensure all recruited participants are followed up 12 months post randomisation. This was a small costed extension to cover salaries and estates/infrastructure costs; as there was an underspend on the study (largely on patient and Trial Manager travel), these funds were vired to cover most of the costs.

Amendments to the protocol

Substantial amendments to the trial protocol were submitted after the initial approval, to clarify statistical changes, dosage clarification for diabetic patients, addition of extra participating centres and in light of the COVID-19 crisis, to permit remote visits.

Version 2.0, dated 5 December 2017: an amendment was made to change the organisation name of the Bristol site that was mistakenly incorrect in version 1.0; and the statistical analysis section was revised as per Medicines and Healthcare Products Regulatory Agency (MHRA) request to detail how subjects who drop out of the study will be analysed and the approach to the analysis of the primary outcome was amended (randomisation stratification variable ‘centre’).

Version 3.0, dated 22 March 2018: typographical errors corrected; and dosage clarification for diabetic patients [recommended a minimum dosage of two (2) × 30 minute sessions per day to better reflect the evidence supporting the diabetic patient group and improvement of their symptoms, rather than a minimum of one (1) × 30 minute daily session].

Version 4.0, dated 7 September 2018: addition of three new participating NHS organisations.

Version 5.0, dated 9 September 2019: addition of additional exclusion criterions to document the criteria that have been followed throughout the duration of the trial; and to permit authorised SET centres to recruit non-SET patients to help aid recruitment into this arm of the trial (patients at SET centres will first be offered the opportunity to attend the SET classes, and only if they do not wish to attend these classes will they be offered the opportunity to participate in the trial as a non-SET patient).

Version 6.0, dated 24 March 2020: an amendment to allow 3-month, 6-month and 12-month visits to take place remotely (i.e. over the telephone completely or in combination with postal questionnaires) in the event that the participant is unable to attend in clinic or the site is unable to accommodate the on-site visit. This was particularly important in light of the COVID-19 pandemic where on-site visits may not be possible.

Sponsorship

The trial was sponsored by Imperial College London.

Study management

Participants

All patients aged ≥18 years, with a diagnosis of IC according to the Edinburgh Claudication Questionnaire (ECQ) and ankle–brachial pressure index (ABPI) (or stress test), were eligible to be included in the trial.

Intervention

Participants in both arms were given local standard care, which includes BMT (such as EA, smoking cessation, etc.) and may include a SET programme, dependent on the Trust, in line with NICE guideline CG147. 4 The SET programme is localised to the Trust (and was not standardised in the study protocol). Table 1 includes the full list of SET sites and their specific SET programme. The SET classes usually involve a circuit of specific lower-limb exercises, supervised by a health-care professional.

Patients in the NMES (intervention) arm also received a REVITIVE IX device (Model: RIX Ref: 1379, Software Version: 2.0). The device is a Class IIa active medical device intended for electrical stimulation of the lower leg in healthy individuals. The indications for use are certified under the Medical Devices Directive 93/42/EEC. The components of the REVITIVE IX device can be found in Figure 1.

FIGURE 1.

Parts and controls of REVITIVE IX. C, foot pads; D, light-emitting diode display panel; E, time-setting controls; F, foot pad intensity controls; G, electrode pad intensity controls; H, power button; I, location of accessory and power sockets; J, IsoRocker. Source: reproduced with permission from REVITIVE IX Circulation Booster: User’s Manual. Bracknell, United Kingdom: Actegy Health Ltd; 2016.

The REVITIVE IX device comes with an alternating current / direct current power adaptor and remote control.

The NMES device delivers electrical stimulation to the participant’s feet via a pair of cushioned foot pads, while they are seated. The IsoRocker feature allows the device to rock back and forth, ensuring adequate stimulation of the calf and foot muscles.

The device is intended for home use for one pre-programmed 30-minute session per day, up to no more than six sessions per day. All devices used in the trial were labelled with the wording ‘exclusively for clinical investigation’ as per MHRA request.

Participants in the standard-of-care (control) arm of the trial received a device at their 12-month study visit.

Inclusion criteria

• positive ECQ

• ABPI <0.9 OR positive stress test (fall in ankle pressure >30 mmHg, 40 seconds post 1 minute treadmill at 10% gradient, 4 km/hour)

• able to give informed consent to participate in the trial after reading the patient information documentation

• age ≥18 years.

Exclusion criteria

• severe IC requiring invasive intervention as determined by the treating clinician

• CLI as defined by the European Consensus Document

• comorbid disease prohibiting walking on a treadmill or taking part in SET

• able to walk for longer than 15 minutes on the study treadmill assessment

• have attended SET classes in the previous 6 months

• popliteal entrapment syndrome

• commenced vascular-symptom-specific medication in previous 6 months, for example, naftidrofuryl oxalate, cilostazol

• pregnancy

• any implanted electronic, cardiac or defibrillator device

• acute deep-vein thrombosis

• broken or bleeding skin, including leg ulceration

• peripheral neuropathy

• recent lower-limb injury or lower back pain

• already using a NMES device.

Sample size

Assuming that the mean AWD in the control group is 200 m following the 3-month treatment period27 with a common equal standard deviation of 120 m,28 and anticipating a 10% rate of loss to follow-up, we estimated that 192 participants would be required to have 90% power with a two-sided alpha level of 5% to detect a difference of 60 m in the mean AWD at 3 months between the intervention and the control group.

Randomisation and treatment allocation

Consenting participants were registered on the web-based data-entry system maintained by Oracle Health Sciences InForm™ electronic data capture (EDC) on an Oracle platform. The randomisation was web-based and blocked with random block size 2, 4 and 6 and stratified by centres. Once eligibility was confirmed, randomisation was performed at the local hospital site by the research nurse prior to any study-related assessments being performed. Participants were randomised to one of the two arms of the trial and assigned a pseudo-anonymised study number unique to each subject enrolled on the study.

Blinding

Due to the nature of the intervention, it was unfeasible to blind the research nurse or participant to the study allocation and a sham device was deemed both impractical and difficult to administer, especially with the REVITIVE IX device causing visual ankle movement. Where possible, a blinded assessor carried out the treadmill test independently and the patient was not given a final score to prevent bias. The senior statistician remained blinded throughout the study.

Settings and location

All participants were recruited from the vascular clinics of 11 secondary-care NHS Trusts throughout England: Imperial College Healthcare NHS Trust; Cambridge University Hospitals NHS Foundation Trust; North Bristol NHS Trust; The Newcastle Upon Tyne Hospitals NHS Foundation Trust; Hull and East Yorkshire Hospitals NHS Foundation Trust; Somerset NHS Foundation Trust (formerly Taunton and Somerset NHS Foundation Trust); University Hospital Southampton NHS Foundation Trust; Nottingham University Hospitals NHS Trust; Dorset County Hospital NHS Foundation Trust; St George’s University Hospitals NHS Foundation Trust; The Royal Bournemouth & Christchurch Hospitals NHS Foundation Trust. For a list of participating hospitals see Acknowledgements, Local research teams.

Sites were selected based on their ability to recruit to the trial, the willingness of the principal investigator (PI) to randomise into the trial and their proven track record in research.

Screening and participant identification

Adult patients presenting to vascular outpatient clinics with a diagnosis of IC were screened by the direct health-care team for eligibility at recruiting centres. The study Research Nurse/Coordinator was notified, who then approached the patient with an information leaflet either in person, via the telephone or by post. Patients were given appropriate time to consider enrolment before consenting assessments were performed.

Recruiting sites also displayed posters describing the study at vascular clinics and the study was presented at many multidisciplinary team meetings to promote awareness among staff.

Anonymous screening logs were completed for all participating sites to log the reasons for non-inclusion along with a minimum data set of age, sex and ABPI (with the permission of the patient). These were frequently sent to the Trial Coordinating Centre to continually monitor recruitment.

Informed consent

Patients who expressed interest in the trial after reading the information leaflet were provided with a patient information sheet (PIS) by the study Research Nurse to consider the trial participation. Consent to enter the study was sought from each subject after a full verbal explanation was given. Potential participants were given ample time to consider study enrolment and ask any questions they may have had.

Written informed consent was obtained from each participant at the screening/baseline visit. The PIS and the consent form both refer to the possibility of linking their data with appropriate databases, including Hospital Episode Statistics and the National Vascular Database, as well as long-term follow-up and access to their NHS records for these purposes. With consent, a letter was also sent to the participant’s general practitioner (GP). A copy of the letter was filed in the Investigator Site File (ISF). The original copy of the signed consent form and PIS were filed in the participant’s local research file (source documents) and a copy was given to the participant.

Written informed consent was obtained before the subject was enrolled in the study.

Baseline assessments

Following written informed consent from the participant, baseline data were collected by the Research Nurse/Coordinator using the case report form. Assessments included the following.

Compliance

Compliance to each of the interventions [EA (part of BMT), SET and NMES] was measured separately to determine the complier/non-complier classification. For each of the treatment groups compliance was defined as follows:

• EA: compliant if completed 75% or more of recommended level of EA (75% of minutes performing exercises recommended by centre).

• SET: compliant if attended 50% or more SET sessions held by centre.

• NMES: compliant if completed 75% or more of recommended level of NMES usage.

Adverse events

The Research Nurse/Coordinator collected all adverse events (AEs) during the duration of the study. AEs were followed up according to local practice until the event had stabilised or resolved. All AEs were assessed for causality and expectedness in relation to the device. The site staff collected occurrences of AEs during follow-up visits, either in person or via telephone or hospital notes.

Follow-up

All randomised participants were followed up until completion of the trial, defined as:

• 12 months post randomisation

• withdrawal from the trial

• death.

Participants in both groups were followed up at 3 months (end of treatment phase), 6 months and 12 months post randomisation. Assessments at this time point included:

• treadmill test

• ABPI/peripheral pulse examination

• QoL questionnaires as per baseline (EQ-5D-5L, SF-36 and ICQ)

• duplex ultrasonography (the DU performed at baseline and 3-month follow-up visit only)

• LDF [the LDF assessment was only performed for 3 minutes at rest at the 6- and 12-month follow-up visit (in both groups)]

• review of participant resource-use diary

• collection of device compliance diary (performed at the 3-month follow-up visit only)

• collection of exercise diary (collected at the 3-month follow-up visit and 6-month visit if the patient continued to attend SET classes following the 3-month visit)

• collection of AEs or SAEs

• drug history review.

The 12-month follow-up appointment marked the end of the study participation. Device participants were able to keep the REVITIVE IX device and control participants were given a device to keep at their 12-month visit.

If consented, patients received weekly text message reminders during the treatment phase to remind them to complete their diaries and attend SET or follow EA.

If the participant was unable to attend the follow-up visit in person or the site was unable to accommodate an on-site visit, the visit could take place remotely (i.e. over the telephone completely or in combination with postal questionnaires). Every effort was made to invite participants back for an on-site visit following a remote visit to complete missed physical assessments, where possible. Sites clearly documented the mode of follow-up that took place.

Participant communications

Participants were kept up-to-date on study progress via Facebook (Meta Platforms, Inc., Menlo Park, CA, USA) and Twitter (Twitter, Inc., San Francisco, CA, USA) accounts. A newsletter summarising the main results from the NESIC trial was also sent to non-withdrawn participants.

Statistical methods

A detailed plan for the analysis of the study outcome data is included in the Statistical Analysis Plan (SAP), which was written and signed off before any final data analysis was commenced. The statistical package STATA version 17 was used to conduct all the analyses undertaken.

All the statistical methods used in the analysis were tested to check if the model’s assumptions were met. Normal distribution of variables was checked by visual inspection using Q–Q plots as well as by using the Shapiro–Wilk test, while homoscedasticity was only visually assessed.

All statistical tests were two-tailed with a 5% significance level.

Additional analyses were undertaken for the raw data (of the variables that were transformed to meet normality assumptions, particularly AWD and ICD) to help with the interpretation. However, these results should be interpreted with caution as the assumption of normality is violated.

Study recruitment

Recruitment commenced in March 2018 and ceased at the end of March 2020. In total 200 participants were recruited from 11 study centres. Table 2 shows the total number of participants recruited per centre. Figure 2 shows the trajectory of recruitment over the study period. At trial commencement, the monthly target recruitment was 1–2 participants/months across the eight study centres (24 in total from each site). To help aid recruitment, a further three study centres were activated in November 2018–January 2019, each with a target of 10 participants each.

FIGURE 2.

Recruitment graph.

Participant flow

The Consolidated Standards of Reporting Trials (CONSORT) diagram is shown in Figure 3. There were 1410 patients assessed for eligibility. Of these 1210 participants were excluded; 326 had an ABPI ≥ 0.9, 166 had a comorbid disease prohibiting treadmill assessment and/or attending the SET programme, 163 declined to participate and 159 had severe IC requiring invasive intervention. See Table 3 for further details. In total, 200 participants were randomised (with 10 included with a positive stress test) and 10 were classified as post-randomisation exclusions. The reasons were that the participant completed the baseline treadmill assessment (walked for longer than the permitted 15 minutes) (n = 5); participant did not complete baseline treadmill test (n = 2); participant could not walk at the required speed in the baseline treadmill test (n = 1); participant randomised but withdrew before completing the baseline visit (n = 1) and participant violated exclusion criteria (ECQ) (n = 1). In the treatment (device) group, 98 participants were randomised and included in the study and in the control group, 102 participants were randomised and included in the study.

FIGURE 3.

Consolidated Standards of Reporting Trials diagram of the trial population. The number of participants who had been lost to follow-up and had died by each time point are presented.

Baseline characteristics for the overall population are shown in Table 4. The mean age in the treatment group (NMES + BMT and NMES + BMT + SET) was 68 years and 67 years in the control group (BMT and BMT + SET). In the treatment group, 76% of participants were male and 71% were male in the control group. The mean BMI in the treatment group was 28 kg/m² and 29 kg/m² (both overweight) in the control group; the majority were former smokers (70% in the treatment group and 59% in the control group) and had a medical history of hypertension and hypercholesterolaemia.

Treatment received

Overall, there were 11 patients (5 in the device group and 6 in the control group) who did not receive the allocated treatment. In the intervention group, five participants were randomised using the incorrect list (incorrect allocation from BMT + SET + NMES to BMT + SET), and in the control group five participants were randomised using the incorrect list (incorrect allocation from BMT + SET to BMT), with a further participant having purchased and used a device following randomisation.

Compliance

Compliance to each of the interventions [EA (part of BMT), SET and NMES] was measured separately. Details of compliance can be found in Table 5. Compliance to EA was the lowest (52.1%) but had the highest percentage of missing data (20.5%). Compliance to SET and the device were similar (69.7% and 73.9%, respectively).

Table 5 was used to classify each patient as either complier or not complier by combining the classifications depending on the treatment assigned (BMT + SET, BMT + SET + NMES, BMT and BMT + NMES). There were 42 out of 190 (22.1%) patients with missing information in the general compliance classification.

Table 6 shows that 40 patients (54.1%) complied with the treatment assigned in the treatment arm, while 46 patients (62.2%) complied in the control arm. There was no statistically significant difference in the proportions of compliers between treatment and control as the difference was −8.1 with a 95% CI of −23.9 to 7.7%; p = 0.32.

Follow-up

Post hoc analysis

Using only the compliance rules for SET and NMES, with all patients having BMT, we identified 124 (65.26%) compliers, but only 117 complete cases in the ITT population were used for the analysis. The results of post hoc analysis to compare the AWD between treatments in the compliers group are found in Table 21. The results indicate that there were no clear differences between the treatment arm and the control arm in AWD in the compliance post hoc analysis and the main analyses. Additional analyses for the primary outcome AWD in all the subgroups were performed (see Report Supplementary Material 3, Tables 22–28).

The outputs of the Wilcoxon rank-sum test for the raw data and the t-test for the transformed square-root AWD to compare treatment at baseline for the post hoc analysis can be found in Report Supplementary Material 3, Table 29. The distributions of AWD at baseline for treatment and control are presented in Report Supplementary Material 3, Figure 5.

The descriptive statistics of AWD at baseline are found in Report Supplementary Material 3, Table 30. Tables for the three strati, short, medium and long distances (set at <25%, 25–75% and >75%, respectively), for AWD can be found in Report Supplementary Material 3, Table 31.

For patients with a short baseline AWD, there was no significant difference between the two treatment arms, nor between type of centre (SET vs. non-SET) (see Table 22).

For patients with a medium baseline AWD, there was no clear difference between the two treatment arms, but there was a significant increase in the square root of the AWD at 3 months (3.081 units; 95% CI 1.01 to 5.14; p = 0.004) for a patient from a SET centre compared with a patient from a non-SET centre (see Table 23). Similarly, for AWD raw data for a patient from a SET centre compared to a non-SET centre, we also see an increase at 3 months (120.59 m; 95% CI 44.21 to 196.98; p = 0.002).

For patients with a long baseline AWD, there were significant differences between treatment arms and type of centre. From Table 24, we observe that there was a significant increase in the square root of the AWD at 3 months (2.877 units; 95% CI 0.51 to 5.25; p = 0.019) and in AWD raw data at 3 months (120.55 m; 95% CI 16.03 to 225.06; p = 0.03) for a patient in the device arm compared with a patient in the control arm. Similarly, there was a significant increase in the square root of the AWD at 3 months (4.033 units; 95% CI 1.61 to 6.45; p = 0.002) and in AWD raw data at 3 months (189.96 m; 95% CI 83.25 to 296.67; p < 0.001) for a patient from a SET centre compared with a patient from a non-SET centre.

Serious adverse events

Table 25 shows SAEs for the overall population categorised by treatment received. SAEs (n = 29) were reported in 24 patients, with all events being classified as either not related or unlikely to be related to the study device. The number of SAEs in the treatment arm was 13 and 16 in the control arm. Most of the events required hospitalisation, there were four deaths and the main primary System Organ Class term for the SAE’s was gastrointestinal disorders.

Introduction

This chapter conducts a within-trial analysis to calculate costs and QALYs over the 1-year time horizon of the NESIC trial. The analysis compares the use of the NMES plus local standard care versus local standard care alone in patients with IC in the UK NHS.

Analyses were undertaken in Stata^® 16 (StataCorp LLC, College Station, TX, USA) and reported according to Consolidated Health Economic Evaluation Reporting Standards guidelines. The health economic analysis plan can be viewed here. Engagement of stakeholders and patients is described in another chapter.

No previous economic studies were found that compared treatment (NMES) versus control (no NMES). One previous economic study30 compared SET to advice only and found that SET was more costly (€3407 vs. €2304) and more effective: 0.71 versus 0.67 QALYs, and a cost per QALY of €28,693 per QALY.

Methods

Results

Cost-effectiveness

Tables 28 and 29 show the mean NHS resource use and costs used by the patients in the study over 3 months and 1 year, for patients with complete follow-up data over those time periods. Figure 4 shows these data graphically. At 1 year, the mean difference in costs between the treatment and control groups was £130, with the initial cost of the device being partially offset by fewer inpatient admissions.

TABLE 28 - Resource use and costs – treatment vs. control: at 3 months, complete cases

Treatment group includes BMT + NMES and BMT + SET + NMES.

Control group includes BMT and BMT + SET.

Notes

Units expressed in mean minutes per patient (except device, one unit per patient; productivity losses, days; out of pocket expenses, pounds). Complete cases are those observations with QoL and cost data for baseline and at 3 months.

Item	Treatmenta (N = 75)			Controlb (N = 71)
	Mean units	Unit cost	Total	Mean units	Unit cost	Total	Difference (treatment – control)
Associated with the intervention
Device	1.00	249.99	249.99	0.00	0.00	0.00	249.99
SET	6.87	5.86	40.26	7.53	5.86	44.13	−3.87
EA	1.00	3.50	3.50	1.00	3.50	3.50	0.00
Costs incurred by patients
Inpatient	0.23	6.43	1.48	0.00	0.00	0.00	1.48
Outpatient	0.67	138.00	92.46	0.69	138.00	95.22	−2.76
GP visit	0.55	30.00	16.50	0.48	30.00	14.40	2.10
Nurse visit	0.43	4.20	1.81	0.35	4.20	1.47	0.34
Health-care professional visit	0.19	57.00	10.83	0.11	57.00	6.27	4.56
Total w/o social costs	416.04			165.16			250.88
Social costs per patient
Productivity losses (days per patient)	0	97.36	0	0.0714	97.36	6.95	−6.95
Out-of-pocket expenses (£ per patient)	–	–	1.60	–	–	0.1019	1.4981
Total	417.64			172.21			245.43

TABLE 29 - Resource use and costs – treatment vs. control: at 12 months, complete cases

Treatment group includes BMT + NMES and BMT + SET + NMES.

Control group includes BMT and BMT + SET.

Note

Complete cases are those observations with QoL and cost data for baseline, 3, 6 and 12 months. A total of 146 observations meet this requirement (treatment, 75; control, 71).

Item	Treatmenta (N = 75)			Controlb (N = 71)
	Mean visits	Unit cost	Total	Mean visits	Unit cost	Total	Difference (t-c)
Device	1.00	249.99	249.99	0.00	0.00	0.00	249.99
Inpatient admission	0.84	130.95	110.00	1.94	113.09	219.40	−109.40
Outpatient visit	2.15	138.00	296.24	2.15	138.00	297.38	−1.14
GP visit	1.84	30.00	55.20	1.79	30.00	53.66	1.54
Nurse visit	1.05	4.22	4.42	0.92	4.22	3.85	0.57
Health-care professional	0.56	57.00	31.92	0.69	57.00	39.34	−7.42
SET	6.87	5.86	40.24	7.54	5.86	44.16	−3.92
EA	1.00	3.50	3.50	1.00	3.50	3.50	0.00
Total	791.51			661.29			130.22
Social costs per patient
Productivity losses (days per patient)	1.2282	97.36	119,57	0.7984	97.36	77.73	41,84
Out of pocket expenses (£ per pat.)	–	–	0.44	–	–	2.13	−1.69
Total	911,52			740,86			170,66

FIGURE 4.

Cost comparison, treatment vs. control at 12 months, complete cases.

Primary cost-effectiveness analysis: treatment versus control

Table 30 shows the unadjusted mean cost and QALY in the treatment and control groups, and Table 31 shows the coefficients of the bivariate regression, adjusted for age, gender and rootMWD in the primary cost-effectiveness analysis. This includes multiple imputations of missing values. Table 32 show the extent of missing data at each time point. The estimated incremental cost per QALY is 188/0.0034 = £55,294/QALY. With a cost-effectiveness threshold of £20,000, the probability that the intervention is cost-effective is 35%, or 42% at a threshold of £30,000 (see Figure 5). None of these effects is significant at the 5% significance level. A parametric cost-effectiveness plane can be seen in Figure 6.

TABLE 30 - Treatment vs. control (N = 190): at 12 months

SE, standard error.

Treatment group includes BMT + NMES and BMT + SET + NMES.

Control group includes BMT and BMT + SET.

Note

Includes multiple imputations of missing values.

	Treatmenta			Controlb			Difference	SE	95% CI
	Mean	SE	95% CI	Mean	SE	95% CI
Cost	777.86	110.40	(556.57 to 999.15)	614.59	159.27	(298.32 to 930.87)	163.26	195.59	(−222.98 to 549.52)
QALY	0.6459	0.0214	(0.6032 to 0.6886)	0.6355	0.0182	(0.5992 to 0.6718)	0.0104	0.0270	(−0.043 to 0.0637)

TABLE 31 - Treatment vs. control: seemingly unrelated regressions (N = 190)

SE, standard error.

Treatment group includes BMT + NMES and BMT + SET + NMES.

Note

Includes multiple imputations of missing values.

	Coefficient	SE	p-value
Total costs
Treatmenta	187.77	196.03	0.338
Age (age – mean)	6.69	11.21	0.551
Gender (female)	357.19	229.56	0.120
Root AWD baseline	−0.0005	0.0008	0.523
Constant	553.06	154.81	<0.000
1-year QALY
Treatmenta	0.0034	0.0191	0.859
Age	0.0010	0.0011	0.369
Female	0.0134	0.0112	0.546
Baseline QoL (EQ-5D)	0.6246	0.0522	<0.000
Root AWD baseline	1.24e-07	8.55e-08	0.147
Constant	0.2346	0.0346	<0.000

TABLE 32 - Missing data at each time point

Treatment group includes BMT + NMES and BMT + SET + NMES.

Control group includes BMT and BMT + SET.

	Treatmenta		Controlb
	Resource use	EQ-5D	Resource use	EQ-5D
Baseline
Complete	91	91	97	97
Missing	1	1	1	1
3 Months
Complete	85	86	84	84
Missing	7	6	14	14
6 Months
Complete	79	80	78	80
Missing	13	12	20	18
12 Months
Complete	77	76	76	76
Missing	15	16	22	22

FIGURE 5.

Treatment vs. control: cost-effectiveness acceptability curve.

FIGURE 6.

Cost-effectiveness plane. Confidence ellipses calculated using means and covariances from the seemingly unrelated regression (R package contour and mvtnorm).

Secondary analyses

SET versus no SET

As a secondary analysis, the costs and QALY associated with SET versus no SET were compared (see Table 33). The incremental cost per QALY of SET versus no SET was 301/0.0232 = £12,974 per QALY. With a cost-effectiveness threshold of £20,000, the probability that the intervention is cost-effective is 58%, or 69% at a threshold of £30,000 (see Figure 7).

TABLE 33 - Supervised exercise therapy vs. non-SET: seemingly unrelated regression (N = 190)

Treatment group includes BMT + NMES and BMT + SET + NMES.

Note

Includes multiple imputation of missing values.

	Coefficient	SE	p-value
Total costs
SET (vs. non-SET)	300.96	193.13	0.119
Age (age – mean)	7.84	11.17	0.483
Gender (female)	282.27	227.80	0.216
Root AWD baseline	−0.0004	0.0008	0.623
Constant	496.11	156.43	0.002
1-year QALY
Treatmenta	0.0232	0.0197	0.239
Age	0.0011	0.0011	0.345
Female	0.0081	0.0222	0.345
Baseline QoL (EQ-5D)	0.6290	0.0519	<0.000
Root AWD baseline	1.27e-07	8.47e-08	0.134
Constant	0.2226	0.0352	<0.000

FIGURE 7.

Supervised exercise therapy vs. non-SET: cost-effectiveness acceptability curve.

Discussion

This chapter has conducted a within-trial economic evaluation of treatment (NMES) versus control (no NMES) for patients with IC. The estimated incremental cost per QALY is £55,294/QALY. With a cost-effectiveness threshold of £20,000, the probability that the intervention is cost-effective is 35%. None of these effects is significant at the 5% significance level. Hence NMES is not considered cost-effective at conventional thresholds in the UK.

The analysis has considered several sensitivity and subgroup analyses. A subgroup analysis in centres that offer SET as local standard care found that the cost per QALY was about £25,000/QALY. An analysis that used an alternative scoring system for the EQ-5D found the cost per QALY was about £11,000/QALY. No other secondary analyses changed the main conclusion that NMES is not cost-effective.

Interpretation of results

NESIC is the first multicentre, pragmatic randomised study to investigate the adjuvant benefit of NMES in patients with IC receiving localised standard care. The results show a treatment hierarchy for patient benefit. Patients in the NMES group in combination with SET and BMT had the most improved AWD at 3 months, followed by patients with access to SET and BMT, and lastly those patients who received BMT alone.

The results of this study add to a growing body of evidence that supports the benefit of SET in improving walking distances in patients with IC. 6–8,41 For this reason, NICE recommends 2 hours of supervised exercise per week for 3 months as first-line management of PAD. Despite this evidence and the current recommendations, it has been shown that SET remains an underutilised tool. Therefore, novel approaches that can be used as an adjunct to local available therapy, designed to increase physical activity in patients with PAD, such as NMES, may have an implication in the first-line treatment of IC.

While this study shows that there is no clear difference in AWD at 3 months between those patients who received NMES compared to standard care alone, there was a non-significant trend suggesting an advantage to NMES when used in combination with SET and BMT. This, taken with the previous body of evidence of improved walking distance, suggests this may be an area for further review.

The primary outcome finding contradicts the outcome in the RCT by Babber et al.,21 which found a significant improvement in MWD after using NMES for 30 minutes daily for a 6-week duration when used independently and also as an adjunct to SET. Considering reasons for this discrepancy, it is noted that the previous study did not reach the recruitment target, while in this study we have also hypothesised that there may be reduced compliance with EA when supplied with an NMES device.

Interestingly, the post hoc analysis showed that patients with a longer baseline MWD have a significantly improved AWD with NMES and SET compared to those with shorter baseline MWD. Neither SET nor NMES was significant in improving AWD in patients who walked <100 m at baseline (N = 40). In contrast, both SET and NMES significantly improved AWD at 3 months for patients who could walk further than 340 m at baseline (N = 40). Perhaps, non-invasive therapies are less effective on those with a poor baseline walking distance and this cohort of patients require surgical intervention. It is noted that these numbers are small but warrant an area for further research.

Health-related quality of life

For health-related QoL, the SF-36 showed no evidence of a difference between the two groups over the follow-up time points for the overall population. The EQ-5D-5L scores significantly improved in the NMES group at 3 months compared to the control group but this was not sustained. After the final 12-month follow-up, there was a significant difference in the disease-specific QoL ICQ improvement between the groups, benefitting the control group. This contrasts with the studies conducted by Babber et al.,21 in which both generic and disease-specific QoL scores showed a significant improvement in the proof-of-concept study, and ICQ scores improved significantly in the NMES group of the RCT. This may be due to the differences in the length of follow-up.

Haemodynamic assessments

Haemodynamic assessments were performed to better understand the underlying mechanisms associated with any changes attributable to the NMES device. There were no clear changes in VF, TAMV or blood flux when the device was turned on, suggesting no improvement in arterial flow to the leg. This contradicts the RCT by Babber et al.,21 which found significant increases in VF and TAMV when the device was switched on at baseline and at week 6, although this was not maintained after device cessation.

Compliance

Compliance with intervention is a vital aspect of the successful management of PAD; 69.7% of those patients with access to SET attended a minimum of 50% of the classes. Current data on compliance with SET in patients with PAD are problematic as the definition of compliance differs between studies and the duration of SET programmes varies widely between research trials. A 2016 systematic review by Harwood et al. 42 of 67 trials showed an average of 75.1% of patients reportedly completed a SET programme; however, only one paper defined a minimal attendance required for SET programme completion.

Compliance to the device in this study was 73.9%, which was less than what was observed in the 6-week pilot study (97%) and subsequent RCT (96%). 21 During the course of this study, no participants contacted the investigators for additional support and most reported good tolerability to device use; 87.5% of device users stated the device was ‘very easy’ to use and 63.6% of device questionnaire respondents stated that they could have used the device more frequently.

The NESIC study also collected data on compliance to EA (52.1%) but this had the highest percentage of missing data (20.5%).

A sensitivity analysis, using only the compliance rules for SET and NMES, with all patients having EA, showed no clear differences from the main analysis (including subgroup analyses).

Cost-effectiveness analysis

There was no significant difference in costs or QALYs between NMES and usual care, in any of the cost-effectiveness analyses undertaken. The subgroup analyses show that QALYs are slightly greater in the SET group compared with no SET, but this does not reach statistical significance. The main cost-effectiveness analysis estimated an ICER that was £55,294 per QALY. However, the gain in QALY was very small and statistically insignificant.

There were some secondary and subgroup analyses where, based on mean costs and QALYs, NMES showed ICERs lower than £30,000 per QALY. Adding NMES to patient care if SET is unavailable gave an ICER of £25,801 per QALY. However, this analysis is ad hoc (not included in the study protocol) and non-randomised. The use of the Devlin scoring algorithm for the EQ-5D gave an ICER of £11,440 per QALY. This algorithm has so far not been recommended by NICE. Other secondary analyses did not show cost-effectiveness at usual NICE thresholds.

Given that the RCT has shown no measurable or clinically relevant difference between NMES and usual care at 1 year, there would be little policy relevance in extrapolating this difference over the longer term in a structured decision model. Likewise, there is little policy relevance in estimating the budget impact of a treatment that does demonstrate cost-effectiveness and is unlikely to be implemented in the population represented by the RCT.

Patient and public involvement

Equality, diversity and inclusion

Geographically diverse recruitment sites were selected to act as participating centres.

These included sites located in local authorities with the highest UK deprivation indices and centres with low research activity where patients may not have had access to research participation.

During the NESIC study, we were contacted by patients who found our study online and wished to participate, which improved accessibility for all patients.

Data relating to equality and diversity (age, gender, ethnicity, recruitment site) were collected from study participants at the initial visit. Data were monitored on a monthly basis by the TMG to ensure that the research sample was representative of the IC population. Any factors limiting equality and diversity in recruitment were reviewed and addressed.

Furthermore, the research team was diverse, interdisciplinary, included patient representatives and had substantial expertise in the delivery of large RCTs in vascular disease.

Generalisability

The trial was designed to be as pragmatic as possible in order to maximise the generalisability of any findings. Patients were recruited from 11 large NHS trusts distributed between those that provide SET and those that provide best medical practice only.

Strengths of NESIC

1. This is the first large RCT looking at the adjuvant benefit of NMES in patients with IC.

2. Generalisability of the results across vascular units that provide a supervised exercise programme and those that provide BMT only.

3. Compliance data collected separately for NMES, SET and EA, with clear definitions on what is deemed as compliant.

Limitations of NESIC

1. Absolute walking distance was used as the primary outcome measure. There was a large variability in the baseline AWD in both groups, with right-skewed distribution. We did not stratify by baseline AWD but the analysis was adjusted by baseline AWD.

2. Only 160 patients had analysable primary outcome data due to missing treadmill data at baseline and/or 3 months.

3. Absence of a sham device comparator. This was considered during the design stage but it was deemed very difficult to blind the research team or participant to the study allocation due to the patient setting the stimulation level to a threshold where calf contractions are visible. This may impact the study findings by introducing bias and we hypothesise that there may be reduced compliance with EA when supplied with an NMES device.

Overall conclusions

This multicentre randomised trial demonstrates the clear benefit of SET for PAD. The addition of NMES may have an adjuvant benefit on AWD, particularly in patients with mild IC. From the subgroup analysis we can conclude that SET and NMES are most effective in patients who are able to walk longer distances at baseline.

For secondary outcomes, the NMES device did not show any improvements in haemodynamic measurements when switched on, nor any significant improvements in generic or disease-specific QoL at the end of the follow-up period. Establishing exercise compliance remains challenging in the PAD cohort, in particular for exercise in a patient’s own time.

Implications for health care

Findings from this trial suggest that all IC patients should have access to a SET programme and changes to such programmes may need to be made to encourage and/or retain participants. NMES may be an effective adjunct to SET and in patients with a good baseline walking distance.

Recommendations for research

• Randomised controlled trial of NMES as an adjunct to SET in IC patients stratified by baseline AWD.

• Research to examine the poor patient motivation and adherence to SET.

• Research to evaluate the long-term effectiveness of SET programmes on MWD and secondary outcomes such as QoL and long-term engagement in physical activity.

Trial applicants

Alun Davies, Joseph Shalhoub, Adarsh Babber, Manjit Gohel, Robert Hinchliffe, James Coulston, Bruce Braithwaite, Ian Chetter, David M Epstein, Gerard Stansby and Francesca Fiorentino.

Trial Management Group

The Trial Management Group comprised Professor Alun Davies (as chief investigator), Ms Laura Burgess (as trial manager), Ms Sasha Smith [as trial manager (maternity cover)], Ms Consuelo Nohpal de la Rosa (as statistician), Dr Francesca Fiorentino (as senior statistician) and Ms Natalia Klimowska-Nassar (as Operations Manager).

Imperial Clinical Trials Unit

The following members were part of the wider ICTU trial team: Amanda Bravery, Kayode Disu, Nayan Das, Ayse Depsen, Smita Das (InForm database team); Ginny Picot, Eloise Britten and Jonathan Dao (quality assurance).

Department of Surgery and Cancer, Imperial College

The following members were part of the wider NESIC study team: Francine Heatley and Rebecca Lawton, Trial Managers; Becky Ward, Sponsor; Matt Ryan and Kathy Lewis, Research Managers, Kirti Patel and Tania Palalic, Contracts.

Trial Steering Committee

We would like to thank Professor Andrew Bradbury (Chair, Sampson Gamgee Professor of Vascular Surgery), Professor Jonathan Beard (Consultant Vascular Surgeon), Dr Louise Brown (Medical Statistician) and Mr Sydney Chapple (lay member), who provided invaluable input and advice as the independent lay member over the course of the study.

Data Monitoring Committee

The team would also like to thank the Independent Data Monitoring Committee members: Professor Julie Brittenden (Chair, Professor of Vascular Surgery), Mr Chung Sim Lim (Consultant Vascular Surgeon) and Dr Stephen Gerry (Medical Statistician) for their support and guidance.

Patient and public involvement

Sydney Chapple was involved in the original design during the grant-application stages and was an active member of the Trial Steering Committee throughout the study. Sydney’s involvement is detailed in the PPI lay-person description (see Report Supplementary Material 2).

Data cleaning

Data cleaning was performed by the trial manager and study statistician.

Local research teams

The NESIC team would like to thank the NHS trusts and participating principal investigators and their colleagues for recruiting and monitoring trial participants. These include (in alphabetical order of participating hospitals followed by the local principal investigators and their colleagues):

Addenbrooke’s Hospital, Cambridge – M.S. Gohel, A. Pentelow, P. Shipley-Cribb, R. Elliot, N. Nacorda, R. Ward, D. Read;

Charing Cross Hospital, London – A. H. Davies, J. Shalhoub, T. Lane, L. Bolton, T. V. Le-Magowan, L. Burgess, B. Jones, N. Strevens, A. M. Malagoni, S. Tavares, A. Henry, C. Connelly, J. Smee, R. Toledano, J. Nunag, L. Tarusan, N. Yasmin, C. Carr;

Dorset County Hospital – J. Metcalfe, B. Page, S. Williams, D. Hill, G. Belt, A. Rees, S. Palmer, S. Horton, D. Lovelock;

Freeman Hospital, Newcastle – G. Stansby, N. Parr, M. Catterson, E. Scott, L. Wales, J. McCaslin, M. Clarke, S. Kirkup, D. Amis, A. Robinson, A. Phillipson, S. Covill, V. Wealleans, E. Fairbairn;

Hull Royal Infirmary – I. Chetter, A. Harwood, J. Long, J. Totty, A. Mohamed, T. Wallace, J. Hatfield, P. Cai, S. Pymer, J. Palmer, A. Firth. T. Roe, S. Ibeggazene, L. Andrews;

Musgrove Park Hospital, Taunton – J. Coulston, A. Stewart, K. Roberts, J. Rewbury, S. Mitchell, H. Mills, L. Vickery, C. Adams, S. Shakya, R. Hadley, L. Timewell, C. Williams, J. Kanapathipillai, J. Hutter, F. Goodchild, N. Greig, J. Blackall, K. O’Callaghan, J. Lucas;

Queen’s Medical Centre, Nottingham – B. Braithwaite, R. Simpson, R. Hadley;

Royal Bournemouth General Hospital – D. Rittoo, C. Thomson, L. Vamplew, M. Letts, T. Webb, E. Howe, A. Fraine, J. Kelly, F. Beecham;

Southampton General Hospital – N. Pal, M. Hulse, P. Patel, I. Nordon, S. Smith, F. Smith, H. Yates, C. Boxall, J. Harvey, S. Hammond;

Southmead Hospital, Bristol – R. Hinchliffe, H. Cheshire, K. Harding, S. McIntosh, L. Poole, P. Brock;

St George’s Hospital, London – P. Holt, N. Sachsinger, R. Ingham, J. Budge, J. Pang, P. Ribeiro.

National Institute for Health and Care Research clinical research networks

We would like to thank the clinical research networks (CRNs) for their help and support throughout the study, in particular CRN North West London (lead site).

Contributions of authors

Laura Burgess (https://orcid.org/0000-0001-7491-7557) (Trial Manager) managed and monitored the trial as trial manager, assisted with acquisition of the data and drafted relevant chapters and approved the final version of the report.

Sasha Smith (https://orcid.org/0000-0001-9843-5368) [Trial Manager (maternity cover)] managed and monitored the trial as trial manager, assisted with acquisition of the data and approved the final version of the report.

Adarsh Babber (https://orcid.org/0000-0002-3056-1752) (Specialty Registrar Vascular Surgery and co-applicant) was responsible for the design, conduct, interpretation of analysis and dissemination and final approval.

Joseph Shalhoub (https://orcid.org/0000-0003-1011-7440) (Consultant Vascular Surgeon and co-applicant) was responsible for the design, conduct, supervision of the study, acquisition of the data, interpretation of analysis and dissemination and drafting relevant chapters and final approval.

Francesca Fiorentino (https://orcid.org/0000-0001-9817-6634) (Senior Statistician and co-applicant) was responsible for the design, conduct of the statistical analysis and drafting relevant chapters.

Consuelo Nohpal de la Rosa (https://orcid.org/0000-0001-7688-2538) (Trial Statistician) was responsible for the conduct of the statistical analysis and drafting relevant chapters.

Natalia Klimowska-Nassar (https://orcid.org/0000-0003-3655-7436) (Operations Manager) was responsible for the design, conduct and dissemination and final approval.

David M Epstein (https://orcid.org/0000-0002-2275-0916) (Lecturer, Applied Economics and co-applicant) was responsible for the design, conduct, analysis, dissemination and drafting of the cost-effectiveness chapter and review of the final draft.

Daniel Pérez Troncoso (https://orcid.org/0000-0003-0091-8148) (PhD student, Applied Economics) was responsible for the analysis, dissemination and drafting of the cost-effectiveness chapter and review of the final draft.

Bruce Braithwaite (https://orcid.org/0000-0002-3828-1819) (Consultant Vascular Surgeon and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

Ian Chetter (https://orcid.org/0000-0002-2566-6859) (Chair of Surgery / Honorary Consultant Vascular Surgeon and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

James Coulston (https://orcid.org/0000-0002-9172-739X) (Consultant Vascular Surgeon and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

Manjit Gohel (https://orcid.org/0000-0001-5685-0723) (Consultant Vascular Surgeon and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

Robert Hinchliffe (https://orcid.org/0000-0002-6370-0800) (Professor of Vascular Surgery and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

Gerard Stansby (https://orcid.org/0000-0001-5539-3049) (Professor of Vascular Surgery and co-applicant) was responsible for the design of the study, acquisition of the data and review of the final draft.

Alun H Davies (https://orcid.org/0000-0001-5261-6913) (Professor of Vascular Surgery) was the chief investigator and was responsible for the design, conduct, supervision of the study, interpretation of analysis and dissemination, drafting relevant chapters and coordination of the report including final approval.

Laura Burgess, Alun H Davies, Francesca Fiorentino, Consuelo Nohpal de la Rosa and David M Epstein were responsible for drafting this report, although all authors provided comments on drafts and approved the final version.

Publication

Lawton R, Babber A, Braithwaite B, Burgess L, Burgess LJ, Chetter I, Coulston J, Epstein D, Fiorentino F, Gohel M, Heatley F. A multicenter randomized controlled study to evaluate whether neuromuscular electrical stimulation improves the absolute walking distance in patients with intermittent claudication compared with best available treatment. J Vasc Surg. 2019;69(5):1567–73.

Ethics and research and development approvals

A favourable ethical opinion was given by the National Research Ethics Service Committee London-Surrey on 20 November 2017 (reference number 17/LO/1918).

Study-wide governance review was undertaken by the North-West London Clinical Research Network (CRN) in November 2017. Research and development (R&D) NHS approvals were granted at the original eight participating sites January–June 2018. The study was granted Health Research Authority approval, January 2018. R&D NHS approvals were granted at the additional three participating sites November–December 2018.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review. Data-access requests are handled on a case-by-case basis and will be reviewed by the corresponding author, Trial Management Group and sponsor. A record of all access to data will be maintained by the Imperial College Archive team.

Disclaimers

This report presents independent research. The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the MRC, the EME programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, the EME programme or the Department of Health and Social Care.