Anal fistula plug versus surgeon's preference for surgery for trans-sphincteric anal fistula: the FIAT RCT

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 07/89/01. The contractual start date was in December 2010. The draft report began editorial review in February 2018 and was accepted for publication in November 2018. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

David G Jayne was a member of the Health Technology Assessment (HTA) Surgery Themed Call Board from 2012 to 2013 and a member of the Efficacy and Mechanism Evaluation Strategy Group from 2015 to 2018. John Scholefield was a member of the HTA Surgery Themed Call Board from 2012 to 2013. Claire T Hulme was a member of the HTA Commissioning Board from 2012 to 2017.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Jayne et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

Objectives

The aim of the FIAT was to compare the Surgisis anal fistula plug with standard surgical treatments for high trans-sphincteric anal fistulas. Standard surgical treatments were the surgeon’s preference of advancement flap, cutting seton, fistulotomy and LIFT procedure.

The specific trial objectives were:

• to determine whether or not the use of the Surgisis anal fistula plug, compared with standard surgical techniques, results in an improvement in symptom-specific QoL

• to determine whether or not the use of the Surgisis anal fistula plug, compared with standard surgical techniques, results in a difference in:

  • fistula healing rates

  • complication and reintervention rates

  • faecal incontinence rates

  • cost-effectiveness

  • health economic benefits.

Trial design

The FIAT is a pragmatic, Phase III, multicentre RCT with a health economic evaluation. Patients with a confirmed high trans-sphincteric fistula at risk of incontinence with fistulotomy (involving approximately one-third or more of the external sphincter complex) were randomised between the insertion of the Surgisis anal fistula plug and the surgeon’s preference of advancement flap, cutting seton, fistulotomy and LIFT procedure (Figure 1).

FIGURE 1.

Trial schema. MRI, magnetic resonance imaging.

Participants

Recruitment and randomisation

Interventions

Participants were randomised to receive either the fistula plug or the surgeon’s preference (advancement flap, fistulotomy, cutting seton or LIFT procedure).

The technique for the placement of the fistula plug was standardised in accordance with the manufacturer’s recommendations for best practice, with all participating surgeons attending a mandatory training session followed by preceptorship with the first cases.

Trial procedures and assessments

Following recruitment into the trial, participants underwent a clinical examination and measurement of St Mark’s incontinence scores. The baseline Faecal Incontinence Quality of Life (FIQoL) and the EQ-5D were completed.

Data on fistula classification according to Parks et al. 21 and on the use of antibiotics and bowel preparation at the induction of anaesthesia, along with technical details of the surgical procedures, were collected intraoperatively. The surgeon’s opinion of the usefulness of the baseline MRI scan as a guide to surgery was also recorded.

Information on the use of postoperative analgesia, immediate complications and reinterventions was collected at the time of discharge.

Participants were followed up at 6 weeks, 6 months and 12 months post randomisation. The trial ended once all participants had completed 12-month follow-up.

At each follow-up visit, a physical examination was performed to determine evidence of fistula healing. Data on bleeding, unexplained pain and septic events, as well as any other complication thought to be related to the intervention, were collected. Medicinal and surgical reintervention rates (i.e. intervention required for an ongoing complication) and St Mark’s incontinence scores were collected at each follow-up visit. Additionally, the FIQoL and EQ-5D questionnaires were completed at the same time points.

At 12 months, all patients underwent follow-up MRI.

Data on serious adverse events (SAEs) were collected at all time points.

Outcome measures

The primary outcome measure for the trial was QoL measured using the validated, symptom-specific FIQoL. QoL was assessed at baseline, at 6 weeks and at 6 and 12 months post randomisation.

Symptom-specific QoL was chosen as the primary outcome, rather than fistula healing rates, as it reflects the primary aim of fistula surgery (to produce symptom relief while maintaining anal sphincter function and preserving symptom-specific QoL).

The secondary outcome measures were:

• fistula healing rate at 12 months

• faecal incontinence rates (as measured by St Mark’s incontinence score) at baseline and at 6 and 12 months

• complication rates at 6 weeks and at 6 and 12 months

• rates of reintervention at 6 and 12 months

• generic QoL assessed EQ-5D and visual analogue scale scores at baseline, 6 weeks, 6 months and 12 months.

Fistula healing at 12 months was assessed by clinical examination. To be deemed to have healed, there had to be no visible external opening and no sign of ongoing sepsis or discharge.

Sample size

The primary outcome measure in the FIAT is symptom-specific QoL measured using the FIQoL questionnaire at baseline, 6 weeks, 6 months and 12 months.

It was estimated that a total of 400 patients would need to be recruited in a 1 : 1 ratio (200 patients to the fistula plug group and 200 patients to the surgeon’s preference group) to be able to detect a small to moderate treatment effect [effect size 0.3 standard deviation (SD)] (i.e. a difference in the primary end point between the two arms of the trial). To allow for a 20% non-compliance rate (non-acceptance, loss to follow-up, incomplete data), the aim was to recruit a total of 500 patients.

The choice of the 0.3 SD treatment effect size was pragmatic. An effect size of 0.2 SD is considered small, 0.5 moderate and 1.0 large. 23 Randomisation of 500 patients in total would provide good statistical power (80% at p < 0.05) to detect an effect size of 0.25 SD, high power (78% at p < 0.01) to detect an effect size of 0.3 SD and very high power (97% at p < 0.01) to detect an effect size of 0.4 SD.

Recruitment to the FIAT was slower than anticipated for a variety of reasons. Lack of surgical equipoise and availability of the plug outside the trial contributed to the low recruitment rate, but the main reason was a lower prevalence of eligible patients than expected. The most common reasons for ineligibility were complex fistula and reclassification owing to MRI results; all patients underwent MRI and a higher proportion than anticipated were reclassified. This led to a change in the diagnostic threshold. The assumptions in the original application were based on routine practice and traditional clinical classification at the time; most surgeons were not using MRI. Practice changed during, and potentially as a result of, the trial, and the unintended consequence was that MRI became more commonplace.

In January 2015, it was agreed with the HTA programme that the sample size for the FIAT would be reduced. If 300 patients were randomised (270 plus 10% dropout), then the FIAT would have 69% power to detect a small to moderate (0.3 SD) treatment effect, or 98% power to detect a moderate (0.5 SD) treatment effect (with α = 0.05).

Statistical methods

The primary outcome measure in the FIAT is symptom-specific QoL measured using the FIQoL questionnaire at baseline and at 6 weeks, 6 months and 12 months. The questionnaire comprises 29 multiple-choice questions grouped into four domains: lifestyle, coping/behaviour, depression/self-perception and embarrassment. Data obtained from the questionnaire were converted into scores using the validated method provided by the developers. Longitudinal plots of mean scores at baseline and over time by treatment group were produced for visual presentation of the data. The primary analysis is a comparison of the mean difference in FIQoL scores between the treatment groups from a repeated measures model (a statistically efficient approach that allows all of the follow-up data collated during the trial to be used, which further enhances statistical power). The model incorporates the 6-week, 6-month and 12-month time points. In addition, the baseline score is included as a covariate in the model. Separate models were constructed for each of the four domains of the FIQoL questionnaire. Further models that included a time-by-treatment interaction term were also fitted. Mean differences and 95% confidence intervals (CIs) were reported.

In addition to the primary adjusted intention-to-treat (ITT) analysis, a ‘per-protocol’ analysis was undertaken for the primary outcome as a sensitivity analysis to explore the potential effect of non-adherence to the randomised allocation. Participants were classified with respect to the first intervention they received, rather than the intervention to which they were randomised. Participants who did not have surgery were excluded from the analysis.

Four a priori subgroup analyses were planned for the primary outcome. These subgroups were for the minimisation variables: age at randomisation (< 30, 30–39, 40–49, 50–59, 60–69, ≥ 70 years), ASA grade (P1, P2, P3, P4), planned type of surgery (advancement flap, cutting seton, LIFT procedure, fistulotomy) and presence of extensions (yes, no). A treatment group-by-subgroup interaction parameter was included in the repeated measures model to assess whether or not there were any differences in the treatment effect across the different strata. Mean differences and 95% CIs were reported.

Data regarding fistula healing were recorded at the 6-week, 6-month and 12-month time points. The fistula healing rates in the two treatment groups were compared using a chi-squared test. Relative risks and 95% CIs were reported.

Faecal incontinence was measured using the St Mark’s incontinence score at baseline and at 6 weeks, 6 months and 12 months. St Mark’s incontinence scores were modelled at each time point, including treatment as a covariate. Mean differences and 95% CIs were reported.

Complication data relating to bleeding, unexplained pain and septic events were recorded at discharge (following operative procedure) and at 6 weeks, 6 months and 12 months. Complication data relating to urinary retention were recorded at discharge (following operative procedure) only. Data regarding participants’ need for reintervention were also recorded at the same time points. Reinterventions were classified as medicinal or surgical by the chief investigator of the trial. The overall complication rates and reintervention rates in the two treatment groups were compared at each time point separately using a chi-squared test. Relative risks and 95% CIs were reported. The proportions of participants experiencing each individual type of complication were also presented, as were the proportions of participants receiving medicinal or surgical reintervention. Time to first reintervention and time to first surgical reintervention were both analysed using a Cox regression model. Hazard ratios and 95% CIs were reported in both cases.

General QoL was assessed using the EuroQol-5 Dimensions, three-level version (EQ-5D-3L), questionnaire at baseline and at 6 weeks, 6 months and 12 months. The questionnaire comprises five multiple-choice questions, each with three possible responses, and a visual analogue scale from 0 to 100. Data obtained from the questionnaire were converted into scores using the validated method provided by the developers. Longitudinal plots of mean scores at baseline and over time by treatment group were produced for visual presentation of the data. The scores in the two treatment groups were compared using repeated-measures models in the same manner as described for the primary outcome. Separate models were constructed for the health status score and visual analogue scale.

Subgroup analyses were carried out for fistula healing, faecal incontinence and EQ-5D-3L in the same manner as described for the analysis of subgroups for the primary outcome.

The SAE data were summarised descriptively. The SAE data were also analysed as a dichotomous variable, with each participant classed as either having or not having experienced a SAE. The two treatment groups were compared using a chi-squared test.

Estimates of treatment effects are presented with 95% CIs and p-values are two-tailed with a p-value < 0.05 considered to be statistically significant. No corrections for multiple tests were made. All analyses were carried out using SAS^® software version 9.4 (SAS Institute Inc., Cary, NC, USA) or Stata^® 14 (StataCorp LP, College Station, TX, USA).

Cost-effectiveness analysis

The aim of the economic evaluation was to assess the cost-effectiveness of the Surgisis anal fistula plug compared with surgeon’s preference for the treatment of trans-sphincteric fistula-in-ano.

The evaluation was performed using a UK NHS and Personal Social Services (PSS) perspective. The evaluation estimated the incremental cost-effectiveness ratios (ICERs) of the fistula plug compared with the surgeon’s preference at 12 months. It was planned that the results would be extrapolated using a decision-analytic model to estimate lifetime cost-effectiveness.

Patient and public involvement

The FIAT was developed by the Research and Audit Committee of the Association of Coloproctology of Great Britain and Ireland (ACPGBI) and supported by the ACPGBI membership and Executive Council (which includes a patient liaison group). Two independent patient representatives were involved in the trial from conception as members of the Trial Management Group.

Ethics approval, regulations and trial registration

Ethics approval for the trial was granted by the National Research Ethics Service Committee East Midlands – Derby Research Ethics Committee (reference number 10/H0405/29).

The trial was conducted in accordance with the recommendations guiding physicians in biomedical research involving human subjects, adopted by the 18th World Medical Association General Assembly in Helsinki, Finland, June 1964,29 and its subsequent amendments, the Research Governance Framework for Health and Social Care,30 and the applicable UK statutory instruments including the Data Protection Act 199831 and the International Conference on Harmonisation Guidelines for Good Clinical Practice.

The trial was prospectively registered as ISRCTN78352529.

Screening

A total of 1355 patients were assessed for eligibility. Of these, 731 patients did not meet the eligibility criteria, 100 patients were eligible but were not randomised and 220 patients were still undergoing screening at the time of trial closure.

Reasons for ineligibility were available for 581 (79%) out of the 731 ineligible patients (Figure 2). Of these 581 patients, 437 (75%) presented with ineligible fistula morphology: complex fistula disease (n = 92), extrasphincteric fistulas (n = 3), fistula healed (n = 23), fistula tract too short (n = 18), high supralevator fistula (n = 1), intrasphincteric fistula (n = 43), low fistula (n = 122), non-cryptoglandular fistula (n = 71), other perineal fistula (n = 32), other unspecified fistula reason (n = 2), no evidence of fistula (n = 19), superficial fistula (n = 9) and suprasphincteric fistula (n = 2).

FIGURE 2.

Reasons for ineligibility.

A total of 94 (16%) patients did not meet the eligibility requirements because of the presence of coexistent anorectal pathology (e.g. anal fissure, haemorrhoids, pilonidal sinus).

A total of 33 (6%) patients were ineligible because of their treatment pathway: previously treated with the fistula plug (n = 18), draining seton not required (n = 5), draining seton not yet in place for 6 weeks (n = 1), cutting seton already in place (n = 1), draining seton to remain in situ (n = 1), contraindication to MRI (n = 5) and MRI outside acceptable time frame (n = 2).

Thirteen (2%) patients were not suitable for surgery; one patient was excluded because she was pregnant and three patients were unable to give informed consent.

Of the 100 patients who were eligible but not randomised, 79 did not want to participate in the trial and the remainder indicated a preference for one of the treatment allocations (12 wanted to receive a fistula plug and nine wanted to receive the surgeon’s preferred treatment).

Recruitment

The FIAT opened to recruitment in May 2011, and the first participant was recruited into the trial on 24 May 2011. A total of 304 participants were recruited and randomised, with the last patient entering the trial on 10 March 2016. Participants were split equally between the two randomised treatment allocations: 152 participants were randomised to receive the fistula plug and 152 were randomised to the surgeon’s preference. The 304 participants were recruited from 40 centres, 75% of those open to recruitment. The number of participants recruited at each site ranged from 1 to 32 (see Appendix 2). Recruitment figures by month are shown in Figure 3 and recruitment figures by centre are shown in Table 2. All participants had reached the 12-month follow-up time point by March 2017.

FIGURE 3.

Recruitment by month.

Participant flow

Of the 304 participants randomised into the FIAT, a total of eight (2.6%) withdrew their consent to remain in the trial. These withdrawals occurred at a range of time points through the trial: five participants withdrew consent prior to any trial treatment, one participant withdrew consent post surgery, one participant withdrew consent before the 6-month time point and one participant withdrew consent before the 12-month time point. Three patients were lost to follow-up: two participants prior to the 6-month time point and one participant before the 12-month time point. In the case of participants who withdrew consent or were lost to follow-up, the data collected up to the point of trial exit were used in the analyses. See Figure 4 for the Consolidated Standards of Reporting Trials (CONSORT) flow diagram with further details of participant flow.

FIGURE 4.

Consolidated Standards of Reporting Trials flow diagram. LFU, lost to follow-up.

Data completeness

Compliance with data collection was good (see Appendix 3). Baseline data were complete for 151 (99%) of the 152 participants in the fistula plug group and for 150 (99%) of the 152 participants in the surgeon’s preference group. Baseline radiology, which was mandatory for inclusion in the trial, was carried out in all participants. Follow-up MRI data were available for analysis for a total of 110 (75%) of the 146 of participants in the fistula group and 112 (77%) of the 146 participants in the surgeon’s preference group. Compliance with the collection of postoperative follow-up data was excellent, with complete data available for 99% of the trial population at the operative and postoperative time points. There was a gradual loss of surgical follow-up data with time, but levels remained acceptable (94% at 6 weeks, 88% at 6 months and 85% at 12 months). Similarly, the completeness of data collected from the patient-reported questionnaires (EQ-5D-3L and FIQoL) decreased with the increased length of follow-up (being 87% at 12 months).

Baseline data

The baseline characteristics of recruited participants, overall and by randomisation group, are shown in Table 3. The majority of participants in each group were classified as ASA I (normal, healthy patient) and were aged between 30 and 60 years (mean 45.1 years, range 18–83 years). There were more men (n = 167, 55%) than women (n = 137, 45%). The most frequently selected surgical procedures, prior to randomisation, were the LIFT procedure (n = 116, 38%) and cutting seton (n = 114, 38%), followed by advancement flap (n = 66, 22%) and fistulotomy (n = 8, 2%). There was no difference in comorbidity between the groups, with smokers making up 23% and 25% of the fistula plug and surgeon’s preference groups, respectively.

The overall incidence of baseline incontinence symptoms, as judged by the St Mark’s incontinence score, was low and similar in the two groups [fistula plug, median incidence 4 (interquartile range 1–6); surgeon’s preference, median incidence 4 (interquartile range 2–8)]. A total of 136 (46%) of the 304 participants reported wearing a pad, which might reflect fistula discharge rather than true anal incontinence.

Chronic sepsis was reported in 98 (65%) patients in the fistula plug group and in 84 patients (56%) in the surgeon’s preference group, with the remainder reporting acute or acute on chronic sepsis. Recurrent fistulas were experienced by 30% of patients in each group, with 64 (42%) patients in the fistula plug group and 73 (48%) patients in the surgeon’s preference group undergoing previous fistula surgery. Overall, the number of prior fistula surgeries was similar between the groups (median 2, range 1–13), with the most frequent operations being seton drainage in 121 of 136 (89%) patients and fistulotomy in 23 of 136 (17%) patients.

All fistulas were deemed to be trans-sphincteric at EUA, although data were missing for one participant in the fistula plug group and two participants in the surgeon’s preference group. The morphology of the fistulas at baseline EUA was similar in both groups, with an overall median primary tract length of 3.0 cm (range 1.5–12.0 cm) and secondary tracts present in 36 (12%) patients. All patients underwent insertion of a draining seton, with the exception of one participant in the fistula plug group and two participants in the surgeon’s preference group for whom data were missing.

In around one-third of participants, gadolinium and oedema MRI sequences were obtained, with similar practices observed at baseline and on follow-up assessment. Baseline MRI characterised the fistula morphology as trans-sphincteric in 132 out of 152 (87%) participants in the fistula plug group and in 138 out of 152 (91%) participants in the surgeon’s preference group, with low numbers of intersphincteric (overall 26, 9%) and superficial (overall 4, 1%) fistulas reported. Secondary extensions on MRI were reported in 25% of cases and at EUA in 12% of cases, with similar numbers of secondary extensions reported in the two groups (overall median 1, range 1–3).

The majority of surgeons (overall 255/304, 84%) only reviewed the MRI report prior to surgery, with only around half actually reviewing the images. Approximately half of surgeons felt that MRI was a useful guide to surgery, with the other half reporting little or no benefit. Review of the MRI findings was reported to alter the surgical approach to a major degree in 12 (4%) of 304 cases and to a minor degree in 37 (12%) of 304 cases.

Compliance with randomisation allocation

Of the 152 participants randomised to the fistula plug group, 135 received their randomised allocation. In 15 participants (9.9%), treatment was non-compliant with the allocated treatment (five received a cutting seton, six received fistulotomy, one received LIFT and three did not receive any surgery). Two participants withdrew from the trial prior to surgery and their treatment is not known.

Of the 152 participants randomised to the surgeon’s preference group, 143 received some form of surgical intervention. Six participants (3.9%) were known to have been non-compliant with their randomised treatment allocation (four received a fistula plug and two did not receive any surgery). Three participants withdrew from the trial prior to surgery and their treatment is not known. At randomisation, all investigators had to indicate which of the four surgeon’s preference options (advancement flap, fistulotomy, LIFT or cutting seton) would be performed if the participant were to be randomised to the surgeon’s preference group. In total, 85% of participants received the type of surgery that was planned (Table 4). Only two fistulotomies were planned, but 13 were carried out.

Overall compliance with randomised treatment was 91.4% (278/304).

Primary end point: quality of life

The primary objective of the FIAT was to compare the fistula plug with standard treatments for high trans-sphincteric anal fistulas in terms of QoL. Symptom-specific QoL was chosen rather than fistula healing rates because it reflects the primary aim of fistula surgery: to provide symptom relief while maintaining anal sphincter function and preserving symptom-specific QoL.

The validated questionnaire used in the FIAT to assess symptom-specific QoL was the FIQoL. Data from this questionnaire were collected at baseline and at 6 weeks, 6 months and 12 months. The questionnaire comprises 29 multiple-choice questions grouped into four domains: lifestyle, coping/behaviour, depression/self-perception and embarrassment. FIQoL domain scores range from 1 to 4, where higher scores indicate higher QoL.

The primary analysis of the FIAT is a comparison of the mean difference in FIQoL scores between Surgisis anal fistula plug and surgeon’s preference from a repeated-measures model incorporating the 6-week, 6-month and 12-month time points, where the baseline score is included as a covariate in the model (Table 5). Separate models have been constructed for each of the four domains of the FIQoL questionnaire. Further models were also fitted, which included a time-by-treatment interaction term to identify any change in treatment over time.

No significant differences were seen in any of the four domains between the fistula plug group and surgeon’s preference group. Models including the treatment-by-time interaction term were also non-significant and, thus, there is no evidence of a change in treatment effect over time.

Figures 5–8 provide longitudinal plots of mean FIQoL scores over time by treatment group for each domain of the FIQoL questionnaire.

FIGURE 5.

Mean FIQoL score: lifestyle domain.

FIGURE 6.

Mean FIQoL score: coping/behaviour domain.

FIGURE 7.

Mean FIQoL score: depression/self-perception domain.

FIGURE 8.

Mean FIQoL score: embarrassment domain.

In addition to the primary-adjusted ITT analysis, a ‘per-protocol’ analysis was undertaken for the primary outcome as a sensitivity analysis to explore the potential effect of non-adherence to the randomised allocation (Table 6 and Appendix 4). Participants were classified with respect to the first intervention they received rather than the intervention to which they were randomised. In total, 155 participants received the surgeon’s preference (143 randomised to the surgeon’s preference; 12 randomised to the fistula plug) and 139 participants received a fistula plug (four randomised to the surgeon’s preference; 135 randomised to the fistula plug). Ten participants did not have surgery and have not been included in this analysis.

Similar to the ITT analysis, the per-protocol analysis failed to show significant differences in any of the domains of FIQoL across the four time points measured (see Appendix 4). A marginal improvement in FIQoL was observed in all domains at 6 weeks following surgery and was maintained until the 12-month follow-up assessment.

Prespecified subgroup analyses of the primary outcome were planned for the four variables for which the randomisation was minimised:

1. age at randomisation (< 30, 30–39, 40–49, 50–59, 60–69, ≥ 70 years)

2. ASA grade (P1, P2, P3, P4)

3. planned type of surgery (advancement flap, cutting seton, LIFT procedure, fistulotomy)

4. presence of extensions (yes, no).

There was no clear evidence to suggest that the treatment effect differed between the different patient subgroups for any of the four FIQoL domains (Tables 7–10).

Secondary outcomes

Reinterventions

Reintervention data were recorded at discharge (following operative procedure) and at 6 weeks, 6 months and 12 months, and each reintervention was classified as either medical (non-operative intervention) or surgical (operative intervention) by the clinical members of the Trial Management Team. The overall analyses of reintervention rates simply consider whether or not the participant received any reintervention at each time point (it is possible that participants could have received multiple reinterventions). In the breakdown of reinterventions, each type is counted separately. However, multiple reinterventions of the same type at the same time point (i.e. medical at 6 weeks) are counted only once. Reintervention data are shown in Table 19.

TABLE 19 - Reinterventions

Values < 1 favour Surgisis anal fistula plug.

Reintervention	Surgisis anal fistula plug (N = 152)	Surgeon’s preference (N = 152)	Risk ratioa (95% CI)	p-value
Postoperative
Complications data available, n	147	144
Reintervention, n (%)	2 (1)	1 (1)	1.96 (0.18 to 21.37)	0.57
Medical, n/N (%)	1/2 (50)	1/1 (100)
Surgical, n/N (%)	1/2 (50)	0/1 (0)
6 weeks
Complications data available, n	142	137
Reintervention, n (%)2	30 (21)	16 (12)	1.81 (1.03 to 3.17)	0.03
Medical, n/N (%)	13/30 (43)	8/16 (50)
Surgical, n/N (%)	18/30 (60)	8/16 (50)
6 months
Complication data available, n	129	129
Reintervention, n (%)2	25 (19)	30 (23)	0.83 (0.52 to 1.34)	0.45
Medical, n/N (%)	3/25 (12)	7/30 (23)
Surgical, n/N (%)	24/25 (96)	24/30 (80)
12 months
Complications data available, n	124	121
Reintervention, n (%)2	28 (23)	27 (22)	1.01 (0.64 to 1.61)	0.96
Medical, n/N (%)	5/28 (18)	2/27 (7)
Surgical, n/N (%)	24/28 (86)	26/27 (96)

Postoperative reinterventions were rare (1% in both groups), highlighting the low immediate morbidity associated with fistula surgery. A significant difference in reinterventions was observed at the 6-week follow-up, being higher in the fistula plug group (n/N = 30/142, 21%) than in the surgeon’s preference group (n/N = 16/137, 12%). This significant difference was lost at 6- and 12-month follow-up, but reinterventions at 12 months were still common in both groups [fistula plug group, n/N = 28/124 (23%) vs. surgeon’s preference group, n/N = 27/121 (22%)]. Medical reinterventions were frequent at 6 weeks, accounting for almost 50% of reinterventions in both arms, but decreased as a proportion of the overall number of reinterventions with progressive follow-up.

Time to first reintervention and time to first surgical reintervention are presented in Figures 9 and 10, respectively. Hazard ratios are computed from a Cox regression model, where values < 1 favour the fistula plug. No significant differences were seen between the treatment groups.

FIGURE 9.

Time to first reintervention. Hazard ratio: fistula plug vs. surgeon’s preference 1.05 (95% CI 0.72 to 1.53).

FIGURE 10.

Time to first surgical reintervention. Hazard ratio: fistula plug vs. surgeon’s preference 1.11 (95% CI 0.75 to 1.66).

Economic evaluation

The economic evaluation was a within-trial analysis and examined the cost-effectiveness of the fistula plug compared with standard surgical techniques based on surgeon’s preference from a UK NHS and PSS perspective.

Health outcomes

Table 28 shows the (unimputed) EQ-5D-3L scores at baseline, 6 weeks, 6 months and 12 months post operation. Both treatment groups showed increasing EQ-5D-3L scores from baseline up to 12 months post operation, with some variation at 6 weeks and 6 months post randomisation in both arms before increasing again.

TABLE 28 - EuroQol-5 Dimensions, three-level version, index scores at baseline and follow-ups, and total QALYs by treatment arm (not imputed)

Time point	Surgisis anal fistula plug (n = 87)	Surgeon’s preference (n = 90)	Difference p-value of t-test
Baseline
Mean (SD)	0.810 (0.238)	0.750 (0.220)	0.0818
Median	0.796	0.796
Minimum–maximum	–0.594 to 1.000	–0.016 to 1.000
6 weeks
Mean (SD)	0.791 (0.222)	0.780 (0.229)	0.7330
Median	0.796	0.796
Minimum–maximum	0.082 to 1.000	–0.181 to 1.000
6 months
Mean (SD)	0.837 (0.210)	0.775 (0.287)	0.0998
Median	0.883	0.796
Minimum–maximum	0.082 to 1.000	–0.239 to 1.000
12 months
Mean (SD)	0.856 (0.195)	0.836 (0.236)	0.5372
Median	1.000	0.924
Minimum–maximum	–0.181 to 1.000	–0.331 to 1.000
Total QALYs
Mean (SD)	0.829 (0.174)	0.790 (0.212)	0.1820
Median	0.852	0.832
Minimum–maximum	0.043 to 1.000	–0.102 to 1.000

On average, the difference between arms was marginal. Independent-sample t-tests indicated that the changes in the EQ-5D-3L score over time were not statistically significant. The average total QALY gain over the 12 months was marginally higher in the fistula plug group (0.829) than in the surgeon’s preference group (0.790), but this difference was not significant (p = 0.182).

Cost-effectiveness results (non-imputed)

Table 29 shows the total costs and EQ-5D-3L-generated QALYs for each of the treatment arms for the complete-case analysis. Differences in QALYs between groups were not significant, with marginal health decrements in the surgeon’s preference group compared with the fistula plug group. The mean total cost was significantly higher for the fistula plug group. The high SD for the deterministic cost estimates reflects the presence of a few outlying individuals who incurred significant health service costs.

TABLE 29 - Total costs and QALYs by treatment arm (deterministic, non-imputed)

QALYs	Surgisis anal fistula plug (n = 87)	Surgeon’s preference (n = 90)
Total QALYs (SD)	0.829 (0.174)	0.790 (0.212)
Total cost, £ (SD)	2738 (1151)	2308 (1228)

Table 30 provides the probabilistic cost-effectiveness results for the non-imputed data, showing the incremental costs and benefits as well as the ICER. The results suggest that the fistula plug is associated with an additional incremental cost of £430 and an incremental QALY gain of 0.039. The ICER shows that the cost per additional QALY gained is £10,993. Although this would be acceptable based on a willingness to pay (WTP) of £20,000 per QALY, this value is subject to much uncertainty. The overall net benefit associated with the fistula plug is positive (£352), suggesting that the fistula plug would be cost-effective over a 12-month time horizon.

TABLE 30 - Probabilistic cost-effectiveness results (non-imputed)

Strategy	Total cost, mean (SD) (£)	Incremental cost, mean (SD) (£)	QALY, mean (SD)	Incremental QALY, mean (SD)	ICER, mean (SD) (£/QALY)	Incremental net benefit, mean (SD) (£)
Surgeon’s preference	2308 (129)		0.790 (0.022)
Surgisis anal fistula plug	2738 (123)	430 (178)	0.829 (0.183)	0.039 (0.029)	10,993 (478,666)	352 (622)

Cost-effectiveness results (imputed)

Considering the imputed data values, the probabilistic cost-effectiveness results for the imputed data are shown in Table 27. There are only minor differences in the mean costs and mean QALYs for the surgeon’s preference group and the fistula plug group compared with the non-imputed data. The ICER is £17,279 per QALY, which is still below the NICE acceptance threshold of £20,000 per QALY, and the incremental net benefit is £71.

However, given the differences in the mean utility values at baseline, as informed by the EQ-5D-3L (see Table 24), the results were adjusted by baseline EQ-5D-3L values. Table 31 shows that when adjusting for baseline EQ-5D-3L values the ICER is £32,400, which is above the NICE acceptance threshold of £20,000 per QALY.

TABLE 31 - Cost-effectiveness results (UK NHS and PSS perspective, probabilistic, imputed)

Strategy	Total cost, mean (SD) (£)	Incremental cost, mean (SD) (£)	QALY, mean (SD)	Incremental QALY, mean (SD)	ICER, mean (SD) (£/QALY)	Incremental net benefit, mean (SD)
Surgeon’s preference	2297 (118)		0.800 (0.021)
Surgisis anal fistula plug	2750 (112)	453 (163)	0.826 (0.018)	0.026 (0.027)	17,279 (1,168,154)	71 (578)
Adjusted baseline EQ-5D-3L					32,400	–168

The uncertainty of the cost-effectiveness estimate for the imputed data with baseline adjustment in the cost-effectiveness plane is shown graphically in Figure 11, using bootstrapping with 1000 iterations. For each bootstrap iteration, a new imputed data set was created, leading to 1000 incremental cost and incremental QALY estimates being produced.

FIGURE 11.

Cost-effectiveness plane showing the incremental cost and QALYs for the fistula plug group compared with the surgeon’s preference group from the bootstrap analysis of the imputed data adjusted for baseline EQ-5D-3L (mean values also shown).

The majority of iterations have a positive incremental cost, which indicates that the fistula plug is very likely to be more costly than the surgeon’s preference. Moreover, the majority of iterations also see a positive incremental QALY estimate, which suggests that the fistula plug is likely to be more effective than surgeon’s preference in terms of QALYs gained. Finally, it is notable that for all 1000 iterations the fistula plug was always more expensive than surgeon’s preference.

The cost-effectiveness acceptability curve showing the probability that the fistula plug is cost-effective is presented in Figure 12 across a range of threshold values of WTP for a single QALY. The probability that the fistula plug is cost-effective is approximately 35–45% across the broader NICE acceptance threshold of £20,000–30,000 per QALY. At WTP thresholds for a QALY above £30,000, the probability that the fistula plug will be cost-effective gradually increases, plateauing at approximately 65%.

FIGURE 12.

Cost-effectiveness acceptability curve showing the probability that the fistula plug will be cost-effective compared with surgeon’s preference across different WTP values for a single QALY, using the bootstrap analysis of the imputed data adjusted for the baseline EQ-5D-3L.

Economic model

Given the lack of secondary sources to inform patient outcomes beyond 12 months, no economic modelling was undertaken.

Limitations

The limitations of the FIAT include the need to reduce the recruitment target because of the lower than anticipated accrual rate. As alluded to above, this was partly due to a high exclusion rate, probably associated with the mandatory use of baseline MRI and perhaps an overestimation in the literature about the true rate of trans-sphincteric fistulas compiled in an era when MRI was not available. Despite a reduction in target recruitment from 500 to 300 patients, the FIAT was still powered to detect a small to moderate difference in the primary outcome (FIQoL). In addition, the original sample size calculation factored in a 20% dropout rate, which did not materialise. Only 7 (2%) out of 304 participants were lost to follow-up and compliance with follow-up data was extremely good. Given the lack of any convincing difference in FIQoL between the fistula plug group and the surgeon’s preference group, it is unlikely that achieving the original sample size of 400 patients would have altered the results. Similarly, the crossover of patients between the treatment arms does not appear to have had a material influence on the results, as demonstrated by the per-protocol analysis.

The FIAT was an unblinded, parallel-group trial. It would obviously not have been possible to blind the surgeons to the treatment allocation. Similarly, it would have been difficult to blind the participants, given that some would have had an anodermal wound (LIFT procedure), whereas others did not (fistula plug and advancement flap) and others would have had a seton protruding from the anal canal (cutting seton). Likewise, blinding of the data collectors would have introduced problems in the collection of treatment-specific information, such as the rate of fistula plug extrusion. Although non-blinding might have introduced an element of bias, perhaps in patient reporting of postoperative pain, it is unlikely to have affected the primary outcome, for which there was no observable difference between the treatment arms.

Strenuous attempts were made to standardise the technique for the placement of the fistula plugs, including surgeon attendance at a mandatory training session, preceptorship with the first three procedures undertaken and video instruction. This was undertaken in an attempt to eliminate any learning curve effect associated with the fistula plug technique, which was probably small given that plug placement is a simple procedure that is well within the skill set of the coloproctologists participating in the FIAT. Standardisation was also considered to be important in reducing the occurrence of early fistula plug extrusion. Despite this, plug extrusion was still reported in 16% of cases of the fistula plug, perhaps indicating an intrinsic problem with the placement of prostheses within the high-pressure anal canal.

A criticism that could be levelled at the FIAT is the small number of patients recruited by several participating sites. This probably reflects the relatively low incidence of trans-sphincteric, cryptoglandular fistula-in-ano, but is mitigated to some degree by the simplicity of the fistula plug technique and the associated short learning curve.

One of the strengths of the FIAT was in the mandatory use of MRI at baseline and the 12-month follow-up, to overcome the recognised inaccuracies in relying solely on clinical examination to characterise fistulas and assess healing. Although we achieved 100% success in obtaining baseline MRI, despite our best efforts we were only able to obtain follow-up imaging for 73% of patients. This is probably still sufficient to obtain reasonable data accuracy and certainly far exceeds, to our knowledge, any previous attempts to use radiological imaging to obtain unequivocal evidence of fistula healing.

Conclusion

The FIAT failed to show a difference in the primary end point, FIQoL, between the fistula plug group and the surgeon’s preference group at the 12-month follow-up. Reassuringly, there was a low rate of early postoperative complications in both groups, indicating that fistula surgery generally carries a low level of morbidity and, therefore, it is justified to perform repeat procedures in an attempt to eradicate the disease. Similarly, it is reassuring to note that incontinence following fistula surgery is generally low, regardless of the surgical technique, and perhaps this should not be used exclusively as a rationale to support the use of the fistula plug. One of the stark highlights of the FIAT is the poor fistula healing rate obtained with all types of fistula surgery. The procedures included in the surgeon’s preference group are those most widely practised by colorectal surgeons in the UK. The results emphasise the inadequacies of current fistula surgery and the need for further research into the underlying pathophysiology of the disease in order to design new, more effective, therapies. Although alternative approaches to fistula eradication continue to be reported in the literature, there is no convincing evidence that they offer any advantage over those evaluated in the FIAT. Importantly, surgeons should learn to manage the expectations of patients suffering with fistula-in-ano, stressing the high failure rates in terms of fistula healing and the need for multiple reinterventions if the aim is to eradicate the disease. The FIAT has provided a health economics evaluation of the fistula plug in comparison with other commonly used surgical techniques. On the basis of the results presented, the fistula plug is more expensive than surgeon’s preference and more effective in terms of QALYs gained. Differences in QALY gains are very small and uncertain in our analysis. Taking account of parameter uncertainty and using some imputation to increase effective sample size, this study suggests that the fistula plug technology is unlikely to be considered a cost-effective use of UK NHS resources.

Recommendations for research

Further research might look at the long-term follow-up of the FIAT cohort to assess whether or not the outcomes change with time. In particular, the lower rates of fistula healing as assessed by MRI, compared with clinical evaluation, might manifest in further fistula recurrences in the future. As previously mentioned, it would also be informative to develop an economic model to examine how patient outcomes beyond the 1-year time horizon impact on the conclusions drawn from the cost-effectiveness analysis. Since the completion of the FIAT, a number of new technologies have been proposed for the treatment of trans-sphincteric fistula and it would be interesting to see if they offer any additional benefit.

Trial Management Group

Professor David G Jayne: chief investigator, applicant, Consultant Surgeon – The Leeds Teaching Hospitals NHS Trust, St James’s University Hospital/Leeds General Infirmary.

Professor John Scholefield: co-applicant, Consultant Surgeon – Nottingham University Hospitals NHS Trust, Queen’s Medical Centre.

Miss Asha Senapati: Consultant Surgeon – Portsmouth Hospitals NHS Trust, Queen Alexandra Hospital.

Dr Damian Tolan: Consultant Radiologist – The Leeds Teaching Hospitals NHS Trust, St James’s University Hospital/Leeds General Infirmary.

Professor Richard Gray: co-applicant, Clinical Trials and Medical Statistics – Nuffield Department of Population Health Medical Sciences Division, University of Oxford.

Dr Claire T Hulme: Health Economics – Academic Unit of Health Economics, Leeds Institute of Health Sciences, University of Leeds.

Dr Laura Magill: Surgical Portfolio Trials Team Leader – BCTU, University of Birmingham.

Dr Kelly Handley: Senior Statistician – BCTU, University of Birmingham.

Mrs Manjinder Kaur: Senior Trial Manager – BCTU, University of Birmingham.

Ms Lucy Prodgers: patient representative.

Ms Claire Roberts: patient representative.

Trial Steering Committee

Mr Simon Ambrose (Chairperson): Consultant Colorectal Surgeon (retired) – The Leeds Teaching Hospitals NHS Trust, St James’s University Hospital/Leeds General Infirmary.

Dr John Spencer (until December 2013): Consultant Radiologist – The Leeds Teaching Hospitals NHS Trust, St James’s University Hospital/Leeds General Infirmary.

Dr James Hampton (from December 2013): Consultant Gastrointestinal Radiologist – Sheffield Teaching Hospitals NHS Foundation Trust, Northern General Hospital.

Dr Ly-Mee Yu: Associate Professor and Deputy Director Academic (Primary Care Trials Unit) – Nuffield Department of Population Health Medical Sciences Division, University of Oxford.

Data Monitoring Committee

Mr Steven Brown (Chairperson): Consultant Surgeon – Sheffield Teaching Hospitals NHS Foundation Trust, Northern General Hospital.

Professor Clive Kay: Chief Executive of Bradford Teaching Hospitals NHS Foundation Trust and Consultant Radiologist – Bradford Teaching Hospitals NHS Foundation Trust, Bradford Royal Infirmary.

Dr Louise Hiller: Principal Research Fellow and Statistician – Warwick Clinical Trials Unit, University of Warwick.

Contributions of authors

Professor David G Jayne (Chief Investigator and Consultant Surgeon) contributed to the conception and design of all aspects of the project, the interpretation of the results and the writing and editing of the report. He also contributed to recruitment, data acquisition and the conduct of the trial.

Professor John Scholefield (Consultant Surgeon) contributed to the conception and design of all aspects of the project. He also contributed to recruitment, data acquisition and the conduct of the trial.

Dr Damian Tolan (Consultant Radiologist) contributed to the conception and design of all aspects of the project, the conduct of the trial, the interpretation of the results and the writing and editing of the report.

Professor Richard Gray (Professor of Statistics) contributed to the conception and design of all aspects of the project, the conduct of the trial, the interpretation of the results and the writing and editing of the report.

Dr Richard Edlin (Health Economist) contributed to the conception and design of all aspects of the project.

Professor Claire T Hulme (Health Economist) conducted and prepared the results of the trial-based cost-effectiveness analysis and contributed to the writing of the report.

Dr Andrew J Sutton (Health Economist) conducted and prepared the results of the trial-based cost-effectiveness analysis and contributed to the writing of the report.

Dr Kelly Handley (Senior Statistician) contributed to the design of the trial, supervised the statistical analyses and the interpretation of the results, and contributed to the writing of the report.

Ms Catherine A Hewitt (Trial Statistician) conducted the statistical analyses, prepared the results and contributed to the writing of the report.

Mrs Manjinder Kaur (Senior Trial Manager) contributed to the conduct and co-ordination of the trial, collation of the data and the interpretation of the results, and the writing and editing of the report.

Dr Laura Magill (Coloproctology/Surgical Portfolio Team Leader) contributed to the conception and design of all aspects of the project, the conduct of the trial, the interpretation of the results and the writing and editing of the report.

Publication

Simpson JA, Banerjea A, Scholefield JH. Management of anal fistula. BMJ 2012;345:e6705.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to available anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). 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, NETSCC, the HTA 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, NETSCC, the HTA programme or the Department of Health and Social Care.

Aneurin Bevan University Health Board (Nevill Hall Hospital)

• Mr Chokkalingam Arun, PI and Lead Surgeon.

• Dr Nick Cross, Lead Radiologist.

• Mr Graham Sturgeon, Lead Research Nurse.

Ashford and St Peter’s Hospitals NHS Foundation Trust (St Peter’s Hospital)

• Mr Philip Bearn, PI and Lead Surgeon.

• Mr Nisar Pasha, Consultant Colorectal Surgeon.

• Dr Allan Irvine, Lead Radiologist.

• Mrs Vitoria Frost, Lead Research Nurse.

Barking, Havering and Redbridge University Hospitals NHS Trust (Queen’s Hospital)

• Mr Joseph Huang, PI and Lead Surgeon.

• Dr Jacques Gutmann, Lead Radiologist.

• Ms Theresa McCluskey, Lead Research Nurse.

Burton Hospitals NHS Foundation Trust (Queen’s Hospital Burton)

• Miss Anna Sverrisdottir, PI and Lead Surgeon (until September 2015).

• Mr Pradeep Thomas, PI and Lead Surgeon (from September 2015).

• Dr Shahzad Khan, Lead Radiologist.

• Ms Gillian Bell, Research Nurse.

• Ms Julie Birch, Research Nurse.

• Ms Clare Mewies, Clinical Trials Co-ordinator.

Calderdale and Huddersfield NHS Foundation Trust (Huddersfield Royal Infirmary)

• Mr Peter Holdsworth, PI and Lead Surgeon.

• Dr Sarah Gurney, Lead Radiologist.

Cardiff & Vale University Health Board (Llandough University Hospital/University Hospital of Wales)

• Mr Michael Davies, PI and Lead Surgeon.

• Dr Robert Bleehen, Lead Radiologist.

• Ms Linda Hazel, Research Nurse.

• Mr Matthew Williams, Research Nurse.

• Ms Micki Palmer, Research Nurse.

Central Manchester University Hospitals NHS Foundation Trust (Manchester Royal Infirmary)

• Mr Finlay Curran, PI and Lead Surgeon.

• Professor Jim Hill, Consultant Surgeon.

• Dr Sarah O Shea, Lead Radiologist.

• Ms Glaxy Gray, Lead Research Nurse.

Chesterfield Royal Hospital NHS Foundation Trust (Chesterfield Royal Hospital)

• Mr Robin Gupta, PI and Lead Surgeon.

• Mr Talalakukoppa Amarnath, Consultant Surgeon.

• Mr Harjeet Narula, Consultant Surgeon.

• Dr Heather Harris, Lead Radiologist.

• Ms Julie Toms, Clinical Research Practitioner.

Croydon Health Services NHS Trust (Croydon University Hospital)

• Mr Muti Abulafi, PI and Lead Surgeon.

• Dr Helena Blake, Lead Radiologist.

Dorset County Hospital NHS Foundation Trust (Dorset County Hospital)

• Mr Michael Lamparelli, PI and Lead Surgeon.

• Dr Peter Taylor, Lead Radiologist.

• Ms Arabis Oglesby, Research Nurse.

• Ms Karen Hogben, Research Nurse.

George Eliot Hospital NHS Trust (George Eliot Hospital)

• Mr Kalimuthu Marimuthu, PI and Lead Surgeon.

• Dr Amitabh Palit, Lead Radiologist.

• Ms Rosemary Musanhu, Lead Research Nurse.

• Ms Emma Brannan, Clinical Trials Co-ordinator.

Gloucestershire Hospitals NHS Foundation Trust (Cheltenham General Hospital)

• Mr James Wheeler, PI and Lead Surgeon.

• Dr Richard Hopkins, Lead Radiologist.

Heart of England NHS Foundation Trust (Birmingham Heartlands Hospital/Good Hope Hospital)

• Mr Haney Youssef, PI and Lead Surgeon.

• Mr David McArthur, Consultant Surgeon.

• Dr Mark Goldstein, Lead Radiologist.

• Ms Linda Webber, Lead Research Nurse.

Homerton University Hospital NHS Foundation Trust (Homerton University Hospital)

• Ms Tamzin Cuming, PI and Lead Surgeon.

• Dr Peter Boavida, Lead Radiologist.

• Ms Sophia Hans, Research Assistant.

Imperial College Healthcare NHS Trust (Charing Cross Hospital/St Mary’s Hospital)

• Mr Gordon Buchanan, PI and Lead Surgeon (until October 2015).

• Mr George Reese, PI and Lead Surgeon (from October 2015).

• Mr Peter Dawson, Lead Radiologist.

• Ms Melloney Allnutt, Research Nurse.

• Ms Gillian Hornzee, Research Nurse.

• Ms Byiravey Pathmanathan, Research Nurse.

Mid Essex Hospital Services NHS Trust (Broomfield Hospital)

• Mr Toby Hammond, PI and Lead Surgeon.

• Dr Peng Lee, Lead Radiologist.

• Ms Joanne Topliffe, Lead Research Nurse.

The Royal Wolverhampton NHS Trust (New Cross Hospital)

• Mr Graham Williams, PI and Lead Surgeon.

• Dr Peter Li, Lead Radiologist.

• Ms Stella Metherell, Lead Research Nurse.

NHS Highland (Raigmore Hospital)

• Professor Angus Watson, PI and Lead Surgeon.

• Mr Michael Lim, Consultant Surgeon.

• Mr Kenneth Walker, Consultant Surgeon.

• Mr James Docherty, Consultant Surgeon.

• Dr Jason Walker, Lead Radiologist.

• Ms Kathleen Macleod, Lead Research Nurse.

Norfolk and Norwich University Hospitals NHS Foundation Trust (Norfolk and Norwich University Hospital)

• Mr Christopher Speakman, PI and Lead Surgeon.

• Dr Stuart William, Lead Radiologist.

• Ms Georgina Glister, Lead Research Nurse.

Nottingham University Hospitals NHS Trust (Queen’s Medical Centre)

• Professor John Scholefield, PI and Lead Surgeon.

• Mr Mike Robinson, Consultant Surgeon.

• Mr Ayan Banerjea, Consultant Surgeon.

• Dr William Dunn, Lead Radiologist.

• Ms Mandy Eyre, Lead Research Nurse.

Oxford University Hospitals NHS Foundation Trust (Churchill Hospital/John Radcliffe Hospital)

• Mr Christopher Cunningham, PI and Lead Surgeon.

• Dr Andrew Slater, Lead Radiologist.

Poole Hospital NHS Foundation Trust (Poole Hospital)

• Mr Guy Nash, PI and Lead Surgeon.

• Dr David Tarver, Lead Radiologist.

Portsmouth Hospitals NHS Trust (Queen Alexandra Hospital)

• Miss Asha Senapti, PI and Lead Surgeon.

• Dr Anthony Higginson, Lead Radiologist.

• Ms Elizabeth Hawes, Research Nurse.

• Ms Sheeba Babu, Research Nurse.

• Ms Karen Flashman, Data Manager.

Royal United Hospitals Bath NHS Foundation Trust (Royal United Hospitals Bath)

• Mr Michael Williamson, PI and Lead Surgeon.

• Dr Andrea Phillips, Lead Radiologist.

• Ms Joyce Katebe, Research Nurse.

• Ms Dawne Chandler, Research Nurse.

Sandwell and West Birmingham Hospitals NHS Trust (Sandwell General Hospital)

• Miss Kathryn Gill, PI and Lead Surgeon.

• Dr Rosamund Donovan, Lead Radiologist.

• Ms Julie Colley, Lead Research Nurse.

• Ms Elzbieta Zulueta, Surgical Care Practitioner Colorectal/General Surgery.

South Tees Hospitals NHS Foundation Trust (The James Cook University Hospital)

• Mr Douglas Atkin, PI and Lead Surgeon.

• Dr Sumeet Miranda, Lead Radiologist.

Southend University Hospital NHS Foundation Trust (Southend Hospital)

• Mr Bandipalyam Praveen, PI and Lead Surgeon.

• Mr Manoj Jacob, Speciality Doctor in Colorectal Surgery.

• Dr Saman Perera, Lead Radiologist.

• Ms Sharon Tysoe, Lead Research Nurse.

Southport and Ormskirk Hospital NHS Trust (Southport and Formby District General Hospital)

• Dr Dimitri Artioukh, PI and Lead Surgeon.

• Dr Apam Chiphang, Lead Radiologist.

• Ms Anna Morris, Research Nurse.

• Ms Dawn Baker, Research Nurse.

St Helens and Knowsley Teaching Hospitals NHS Trust (Whiston Hospital)

• Mr Raj Rajaganeshan, PI and Lead Surgeon.

• Dr Kirsty Slaven, Lead Radiologist.

Taunton & Somerset NHS Foundation Trust (Musgrove Park Hospital)

• Ms Louise Hunt, PI and Lead Surgeon.

• Mr Paul Mackey, Consultant Surgeon.

• Dr Paul Burn, Lead Radiologist.

• Ms Jayne Foot, Lead Research Nurse.

Hillingdon Hospitals NHS Foundation Trust (Hillingdon Hospital)

• Mr Yasser Mohsen, PI and Lead Surgeon.

• Mr Alistair Myers, Consultant Surgeon.

• Dr Ziad Meer, Lead Radiologist.

• Miss Alex Diaz, Research Nurse.

• Ms Mariam Nasseri, Research Nurse.

• Ms Melinda Holden, Research Nurse.

Ipswich Hospital NHS Trust (Ipswich Hospital)

• Mr James Pitt, PI and Lead Surgeon.

• Dr Simon Smith, Lead Radiologist.

• Ms Claire Swann, Lead Research Nurse.

Leeds Teaching Hospitals NHS Trust (St James’s University Hospital/Leeds General Hospital)

• Professor David Jayne, PI and Lead Surgeon.

• Dr Damian Tolan, Lead Radiologist.

• Ms Catherine Moriarty, Lead Research Nurse.

The Mid Yorkshire Hospitals NHS Trust (Pinderfields General Hospital/Dewsbury and District Hospital)

• Mr Chris Macklin, PI and Lead Surgeon.

• Dr Katherine Naik, Lead Radiologist.

• Ms Stephanie Lupton, Lead Research Nurse.

Royal Liverpool and Broadgreen University Hospitals NHS Trust (The Royal Liverpool University Hospital)

• Mr Paul Rooney, PI and Lead Surgeon.

• Dr Mark Hughes, Lead Radiologist.

• Dr Priya Healey, Lead Radiologist.

University Hospitals Birmingham NHS Foundation Trust (Queen Elizabeth Hospital Birmingham)

• Mr Nigel Suggett, PI and Lead Surgeon.

• Professor Dion Morton, Consultant Surgeon.

• Dr Deborah Tattersall, Lead Radiologist.

• Dr Manijeh Ghodds, Data Contact.

University Hospitals Bristol NHS Foundation Trust (Bristol Royal Infirmary)

• Mr Michael Thomas, PI and Lead Surgeon.

• Dr Huw Roach, Lead Radiologist.

• Dr Mark Callaway, Lead Radiologist.

• Ms Rebecca Houlihan, Research Nurse.

• Ms Karen Bobruk, Research Nurse.

• Ms Catherine Phillpott, Research Nurse.

University Hospitals of Leicester NHS Trust (Leicester General Hospital)

• Mr Baljit Singh, PI and Lead Surgeon.

• Dr Ratan Verma, Lead Radiologist.

Wirral University Teaching Hospital NHS Foundation Trust (Arrowe Park Hospital)

• Mr Liviu Titu, PI and Lead Surgeon.

• Dr Nicholas Day, Lead Radiologist.

• Ms Helyn Evans, Lead Research Nurse.

Yeovil District Hospital NHS Trust (Yeovil District Hospital)

• Mr Nader Francis, PI and Lead Surgeon.

• Dr Charles Hopkins, Lead Radiologist.

• Mrs Joanna Alison, Research Nurse.

• Ms Alison Lewis, Research Nurse.