Systematic review

Systematic review of tibial nerve stimulation to treat faecal incontinence E. J. Horrocks1 , N. Thin1 , M. A. Thaha1 , S. J. C. Taylor2 , C. Norton3 and C. H. Knowles1 1

National Centre for Bowel Research and Surgical Innovation and 2 Centre for Primary Care and Public Health, Blizard Institute, Queen Mary University London, and 3 Florence Nightingale School of Nursing and Midwifery, King’s College London, London, UK Correspondence to: Miss E. J. Horrocks, National Centre for Bowel Research and Surgical Innovation, 1st Floor, Abernethy Building, 2 Newark Street, London E1 2AT, UK (e-mail: [email protected])

Background: Two forms of tibial nerve stimulation are used to treat faecal incontinence (FI):

percutaneous (PTNS) and transcutaneous (TTNS) tibial nerve stimulation. This article critically appraises the literature on both procedures. Methods: A systematic review was performed adhering to the PRISMA framework. A comprehensive literature search was conducted, with systematic methodological quality assessment and data extraction. Summary measures for individual outcome variables are reported. Results: Twelve articles met eligibility criteria; six related to PTNS, five to TTNS, and one to both procedures. These included ten case series and two randomized clinical trials (RCTs). Case series were evaluated using the National Institute for Health and Care Excellence quality assessment for case series, scoring 3–6 of 8. RCTs were evaluated using the Jadad score, scoring 4 of a possible 5 marks, and the Cochrane Collaboration bias assessment tool. From one RCT and case series reports, the success rate of PTNS, based on the proportion of patients who achieved a reduction in weekly FI episodes of at least 50 per cent, was 63–82 per cent, and that of TTNS was 0–45 per cent. In an RCT of TTNS versus sham, no patient had a reduction in weekly FI episodes of 50 per cent or more, whereas in an RCT of PTNS versus TTNS versus sham, 82 per cent of patients undergoing PTNS, 45 per cent of those having TTNS, and 13 per cent of patients in the sham group had treatment success. Conclusion: PTNS and TTNS result in significant improvements in some outcome measures; however, TTNS was not superior to sham stimulation in a large, adequately powered, RCT. As no adequate RCT of PTNS versus sham has been conducted, conclusions cannot be drawn regarding this treatment. Paper accepted 7 November 2013 Published online 20 January 2014 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.9391

Introduction

Faecal incontinence (FI) is a significant public health problem, with a prevalence in the UK estimated to be 1–15 per cent amongst adults outside care homes, rising to as high as 50 per cent in care homes where FI is overrepresented in those with severe mental and/or physical disability1,2 . Further, only 15–45 per cent of those suffering from FI consult medical services, owing to embarrassment and lack of knowledge about potential treatments3 – 5 . FI has a significant impact on quality of life (QoL), causing social and psychological disability6,7 , and often leads to people suffering from stigmatization and social exclusion8 – 10 . The attendant socioeconomic burden of FI is high11,12 . Management of FI is challenging owing to a combination of its high prevalence, aetiological heterogeneity and lack  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

of widespread expertise. Management usually involves a stepwise approach, beginning with more conservative strategies and moving on to appropriately tailored medications, nurse-led bowel retraining programmes or focused biofeedback, and psychosocial support. A combination of these treatments often improves symptoms, although they are not universally successful13,14 . Patients with refractory symptoms may have the option of a surgical solution, such as sphincter repair, artificial sphincter, dynamic graciloplasty or, finally, a permanent stoma. Surgical procedures are invasive and have, at best, variable success rates with significant risk of morbidity15 – 18 . Neuromodulation is a relatively new treatment modality for FI that is based on recruitment of residual anorectal neuromuscular function pertinent to continence by electrical stimulation of the peripheral nerve supply, BJS 2014; 101: 457–468

458

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

without the need for potentially hazardous surgery to the anus itself. Sacral nerve stimulation (SNS) employs direct electrical stimulation of the sacral nerve roots (mainly S3 nerve root) and is a safe, effective treatment for FI regardless of aetiology19 – 25 . Despite largely favourable data, SNS does not provide a cure for all patients, is not without risk of morbidity26 and, although cost-effective compared with other surgical options27 , does have high associated costs28 . Tibial nerve stimulation (TNS) is a newer alternative. The tibial nerve contains afferent and efferent fibres originating from the fourth and fifth lumbar nerves and the first, second and third sacral nerves. Thus, stimulation of the tibial nerve may lead to changes in anorectal neuromuscular function similar to those observed with SNS (owing to shared sacral root effects), but without the need for a permanent, surgically implanted, device. First described in 1983 by McGuire and colleagues29 for urinary incontinence using a transcutaneous electrode, the method was adjusted in 1999 by Stoller30 by use of a percutaneous needle with a ground electrode on the ipsilateral extremity. In 2003, Shafik and co-workers31 proposed using percutaneous tibial nerve stimulation (PTNS) for FI. Two methods of outpatient delivery of TNS have been described in the literature. The first method of delivery is PTNS, in which a needle is placed near to the tibial nerve in the ankle, just above the medial malleolus, and electrical stimulation is applied via this needle. Two sets of equipment for delivery of PTNS are described: the Stoller afferent nerve stimulator (SANS; UroSurge, Coralville, Iowa, USA), which consists of a stimulator (a 9-V AC monopolar generator with stimulation range of 0·5–10 mA, pulse frequency 20 Hz, pulse width 200 µs) attached with a connecting lead to a 34-G needle and a neutral surface electrode; and the Urgent PC neuromodulation system (Uroplasty Limited, Manchester, UK), which consists of a stimulator (adjustable current setting of 0–9 mA in preset 0·5-mA increments, fixed pulse frequency 20 Hz, pulse width 200 µs) that provides visual and auditory feedback. Twelve sessions of treatment are usually given (over 6 or 12 weeks), with each session lasting for 30 min. After the 12 initial treatments, some patients need ‘top-up’ sessions to sustain symptom relief. Side-effects include occasional bleeding and tenderness at the site of needle insertion32,33 . The second method of delivery is transcutaneous electrical nerve stimulation (TENS), in which two pad electrodes are placed over the tibial nerve just above the ankle. There are many TENS machines on the market, but stimulation parameters are set to a pulse frequency of 10 Hz and a pulse width of 200 µs, with stimulation current of 0–30 mA.

Each study in this review used a different treatment protocol in terms of treatment frequency and duration, although each session was 20–30 min in length. This second method will be referred to as transcutaneous TNS (TTNS). Two review articles34,35 on TNS for FI have described published studies in some detail; however, neither summarized results to give an idea of overall effectiveness. Furthermore, two randomized studies36,37 have been published since these reviews. The aim of the present review is to provide a comprehensive and systematic overview of PTNS and TTNS in the treatment of FI. The two methods of TNS described above are addressed separately.

 2014 BJS Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

Methods

A systematic review was performed, using the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) framework38 to minimize risk of bias. The authors developed the protocol for review, detailing prespecified methods of the analysis and eligibility for the review. The following criteria were used.

Inclusion criteria Studies of any design, including case series, case–control and randomized clinical trial (RCT) designs, reporting TNS for FI from January 2003 to February 2013 were eligible for inclusion. The study must have provided data for at least ten patients treated with PTNS or TTNS. Only reports that clearly reported at least one of the selected review outcome measures (baseline and postintervention) and a clear follow-up period were eligible for inclusion. No exclusions were placed on study centre or patients, in terms of age, sex, ethnicity or aetiology of FI. All eligible studies required a definitive intervention by PTNS or TTNS for FI; reports of PTNS or TTNS for primarily urinary symptoms, or mixed symptoms, were excluded.

Outcomes The primary outcome prespecified was the success rate of therapy, based on an improvement of at least 50 per cent in the number of FI episodes. Secondary outcomes were: reduction in weekly FI episodes; cure rates of the treatment (100 per cent reduction in episodes); improvement in the Cleveland Clinic Incontinence Score (CCIS) or similar (including St Mark’s incontinence score)39 ; improvements in QoL measures (generic and condition-specific); and improvements in any other outcomes reported by individual studies. BJS 2014; 101: 457–468

Tibial nerve stimulation in faecal incontinence

Methods for quality appraisal Critical appraisal of the studies was considered by study type and data collected (Tables S1–S3, supporting information) and potential sources of bias were considered. Case series, which comprised most of the PTNS and TTNS studies, were assessed using the National Institute for Health and Care Excellence (NICE) ‘Quality Assessment for Case Series’ system40 , which assesses characteristics of methodology, outcomes and interpretation from a possible score of 8. The two randomized trials were evaluated using the Jadad score41 , which awards points for randomization method, blinding and account of all patients. The score is out of a possible 5 points, with a score of above 3 indicating good quality. The RCTs were also evaluated using the Cochrane Collaboration tool for assessing risk of bias in randomized trials42 , which evaluates six specific domains and reports on whether there is a high, low or unclear risk of bias in each area.

459

from the relevant institutions were contacted to confirm or refute any repetition of results.

Summary measures and analysis Formal data synthesis was not possible owing to heterogeneity between study designs and outcome measures, and the absence of any genuine comparator for nearly all studies. Data synthesis was attempted and all authors were contacted to request receipt of the raw data set. Replies were received from only two authors: one no longer had access to the original data set, and the other was reluctant to share the raw data. Further data synthesis was therefore not possible. Summary measures for individual outcome variables across broadly homogeneous studies were limited to descriptions of percentages, medians and ranges. Results

Search Two authors carried out a comprehensive search of the literature on 10 February 2013 using PubMed, MEDLINE, Embase and Evidence-Based Medicine reviews (including the Cochrane database of systematic reviews and the Cochrane central register of controlled trials). Search terms used were ‘tibial nerve stimulation faecal incontinence’ (‘‘tibial nerve’’[MeSH Terms] OR (‘‘tibial’’[All Fields] AND ‘‘nerve’’[All Fields]) OR ‘‘tibial nerve’’[All Fields]) AND stimulation[All Fields] AND (‘‘faecal incontinence’’[All Fields] OR ‘‘fecal incontinence’’[MeSH Terms] OR (‘‘fecal’’[All Fields] AND ‘‘incontinence’’[All Fields]) OR ‘‘fecal incontinence’’[All Fields]). Full-text copies of all studies deemed to be potentially relevant were obtained and assessed for inclusion. Where papers cited other potentially important references, these were also assessed. Systematic reviews, RCTs and case series of patients with FI who had undergone TNS were searched for relevant data sets. The reviewers were not blinded to the names of studies, authors, institutions or publications. Search results were cross-referenced with bibliographies of relevant papers.

The initial electronic search for PTNS and TTNS revealed 28 relevant citations, of which 14 reports of 12 studies remained when eligibility criteria were applied (Fig. 1). Studies were published between 2003 and 2013, and included a total of 375 patients (range 10–88) undergoing active treatment for FI of various aetiologies. Of these, six studied PTNS, five TTNS, and one both PTNS and TTNS. Results are presented as those from randomized studies, those from PTNS studies (including the PTNS arm of randomized studies), and those from TTNS studies (including the TTNS arm of randomized studies). Potentially relevant articles identified and abstracts screened n = 28

Full-text articles reviewed n = 14

Excluded (exact duplicate data series) n = 2

Study selection Eligibility assessment was performed by the same two authors, in an unblinded but standardized manner. Methodological quality of included studies was assessed independently using the PRISMA assessment criteria38 . Disagreement was resolved by the senior author. Exact duplicate data sets generated from the same cohort of patients were excluded. In instances of doubt, authors  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

Excluded n = 14 Full-text article not available n=1 Article not relevant n = 8∗ Review article n = 3 Video demonstration n = 1 Fewer than 10 patients in series n = 1

Articles included in systematic review n = 12

PRISMA diagram for the systematic review. *No percutaneous or transcutaneous tibial nerve stimulation (6 studies), pelvic order disorder (1 study) and chronic pelvic pain (1)

Fig. 1

www.bjs.co.uk

BJS 2014; 101: 457–468

460

Table 1

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

Critical appraisal of randomized clinical trials using the Cochrane Collaboration tool42 for assessing risk of bias in randomized

studies Domain Leroi et al.36 (2012) (France) Sequence generation Allocation concealment Blinding of participants, personnel and outcome assessors Incomplete outcome data Selective outcome reporting Other sources of bias George et al.37 (2013) (UK) Sequence generation Allocation concealment Blinding of participants, personnel and outcome assessors Incomplete outcome data Selective outcome reporting Other sources of bias

Review authors’ judgement of risk of bias

Description

Random number table stratified based on CCIS score Random number table stratified based on CCIS score – unclear whether allocation central, web-based or other Double-blinded – patient and evaluating physician blinded to treatment 13 patients dropped out and therefore not analysed All outcomes reported None

Low Unclear Low High Low

Sealed-envelope randomization Sealed, windowless envelopes used Single-blinded – patient blinded only

Low Low High

No missing outcome data All outcomes reported Paper reported that sham arm was suitable only for comparison with TTNS arm, and not for PTNS arm

Low Low High

CCIS, Cleveland Clinic Incontinence Score; TTNS, transcutaneous tibial nerve stimulation; PTNS, percutaneous tibial nerve stimulation.

Critical appraisal of percutaneous and transcutaneous tibial nerve stimulation case series using the National Institute for Health and Care Excellence Quality Assessment for Case Series form40

Table 2

Reference PTNS series Shafik et al.31 (2003) de la Portilla et al.43 (2009) Govaert et al.44 (2010) Boyle et al.33 (2010) Findlay et al.45 (2010) Hotouras et al.46 (2012) Median (range) TTNS series Queralto et al.47 (2006) Vitton et al.48 (2009) Vitton et al.49 (2010) Eleouet et al.50 (2010) Median (range)

Score

3 4 5 4 4 5 4 (3–5) 3 4 6 5 4·5 (3–6)

PTNS, percutaneous tibial nerve stimulation; TTNS, transcutaneous tibial nerve stimulation.

Quality of studies Randomized studies Using the Jadad score41 , both RCTs scored 4 of a possible 5 marks, indicating good quality. Appraisal of RCTs using the Cochrane Collaboration tool for assessing risk of bias (Table 1) identified one area of potential bias, in the form of incomplete outcome data in the larger RCT comparing TTNS with sham stimulation36 . In the smaller RCT comparing PTNS versus TTNS versus sham37 , two areas  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

of potential bias were identified (blinding and conduct of sham).

Non-randomized studies Critical appraisal of the case series31,33,43 – 50 , using the NICE quality assessment form, is shown in Table 2. All studies had a score between 3 and 6 of 8. The median score for TTNS studies was 4·5 (range 3–6) and that for PTNS studies was 4 (3–5). Results of randomized studies The two randomized studies of PTNS and TTNS included a large multicentre RCT of TTNS versus sham containing 144 patients36 and a smaller single-centre study of PTNS versus TTNS versus sham containing 30 patients37 . The outcomes of these RCTs are summarized in Tables 3 and 4 (full version of Table 3 is supplied as Table S4, supporting information); further details of treatment protocols are given in Table S1 (supporting information).

Primary outcome In the TTNS versus sham study36 , no patient in either group had a 50 per cent or greater reduction in weekly FI episodes. In the PTNS versus TTNS versus sham study37 , 82 per cent of patients in the PTNS group, 45 per cent of patients in the TTNS group and 13 per cent of those in the sham group had treatment success based on this outcome (Table 3). This was statistically significant across all groups: there were significantly more www.bjs.co.uk

BJS 2014; 101: 457–468

Tibial nerve stimulation in faecal incontinence

461

Faecal incontinence episodes and Cleveland Clinic Incontinence Score outcomes in randomized trials of percutaneous and transcutaneous tibial nerve stimulation

Table 3

Reference

Follow-up (months from Initial treatment start of period treatment) (months)‡

n*

FI episodes/week#

Median (range)§

CCIS**

P

≥ 50% ≥ 50% improvement in CCIS (% of reduction patients) (%)

Median CCIS§

P

Leroi et al.36 TTNS 73 (68) Sham 71 (63) George et al.37 PTNS 11 (n.r.) TTNS 11 (n.r.) Sham 8 (n.r.)

3 (180) 3 (180)

3 3

1·5 (12) 1·5 (12) 1·5 (12)

1·5 1·5 1·5

1·7 (0–23) to 1·0 (0–14·3) (−0·7) 0·004 2·9 (0–25) to 1·6 (0–23·6) (−1·3) 0·06 8·2(5·2) to 1·8(0·8)(−6·4)† 7·4(5·9) to 5·1(4·2)(−2·3)† 6·5(3·5) to 4·7(3·5)(−1·8)†

n.r. n.r. n.r.

0 0

47¶ 27¶

82 45 13

– – –

11 to 8 (−3) < 0·001 11 to 9 (−2) < 0·001 – – –

– – –

*Values in parentheses are numbers of women. †Data reported as mean(s.d.). ‡Number of initial treatments shown in parentheses; there were no top-up treatments in either study. §From baseline to follow-up; difference is shown in parentheses. ¶More than 30 per cent improvement in Wexner score. FI, faecal incontinence; CCIS, Cleveland Clinic Incontinence Score; TTNS, transcutaneous tibial nerve stimulation; n.r., outcome measured but not reported; PTNS, percutaneous tibial nerve stimulation. #Leroi et al.36 , no significant difference between groups; George et al.37 , P = 0·042 (PTNS versus TTNS and sham). **Leroi et al.36 , no significant difference between groups. Table 4

Other outcomes in randomized trials of percutaneous and transcutaneous tibial nerve stimulation

Reference

Outcome

Results

Leroi et al.36

FIQL score Patient-perceived success

No significant differences between the two groups* No difference in patient-perceived treatment efficacy between groups: TTNS 30 (0–100) per cent versus sham 20 (0–100) per cent success (P = 0·024) Physician-estimated treatment efficacy significantly higher in TTNS group: TTNS 59 per cent versus sham 35 per cent improved (P = 0·01) No significant difference between two groups in anorectal manometry at 3 months compared with baseline All groups demonstrated improvement in incontinence score, but no significant difference between groups (P = 0·201) All groups demonstrated improvement in FIQL score, but no significant difference between groups in any domain All groups demonstrated improvement in SF-36 score, but significant difference between groups only in vitality domain (P = 0·008) No anorectal physiology assessments showed any significant difference between groups

Physician-perceived success Anorectal manometry George et al.37

St Mark’s incontinence score FIQL score SF-36 QoL score Anorectal manometry

*In terms of median relative changes compared with baseline. FIQL, Faecal Incontinence Quality of Life Scale; TTNS, transcutaneous tibial nerve stimulation; SF-36, Short Form 36 (QualityMetric, Lincoln, Rhode Island, USA).

patients with treatment success in the PTNS group than in the TTNS group, and in the TTNS group compared with the sham group (P = 0·035, Fisher’s exact test).

significance within the groups was not reported. Statistical analysis demonstrated a significantly greater reduction in the PTNS group (P = 0·042). CCIS was not collected in the three-arm study.

Secondary clinical outcomes In the TTNS versus sham study36 , there was a significant reduction in median number of weekly FI episodes, from 1·7 (range 0–23) to 1 (0–14·3) in the TTNS group (P = 0·004), and in median CCIS in both TTNS (from 11 to 8) and sham (from 11 to 9) groups (P < 0·001) (Table 3). However, there were no significant differences between the two groups. In the three-arm study37 , there was a reduction in the mean number of FI episodes in the PTNS group (from 8·2 to 1·8), in the TTNS group (from 7·4 to 5·1) and in the sham group (from 6·5 to 4·7). However,

Quality of life In the TTNS versus sham study36 , both TTNS and sham groups had significantly improved scores in all four Faecal Incontinence Quality of Life Scale (FIQL) domains compared with baseline, but there were no significant between-group effects. In the three-arm study37 , all groups demonstrated improvement in the FIQL and Short Form 36 (SF-36; QualtyMetric, Lincoln, Rhode Island, USA) scores; however, there were few statistically significant differences between groups (Table 4).

 2014 BJS Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

BJS 2014; 101: 457–468

462

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

Faecal incontinence episodes and Cleveland Clinic Incontinence Score outcomes in trials of percutaneous tibial nerve stimulation

Table 5

FI episodes/week No. of Follow-up initial weekly (months from treatments start of (weeks) treatment)

Reference

n

Shafik et al.31‡ de la Portilla et al.43

32 16

14 (4) 12 (12)

Govaert et al.44

22

12 (6)

Boyle et al.33 Findlay et al.45 Hotouras et al.46 George et al.37

31 13 88 11

Summary (median (range))

213

Median (range)†

P

CCIS

≥ 50% improvement ≥ 50% in CCIS reduction (%) (% of patients) n.r. 38 44 31 n.r.

Median CCIS†

n.r. n.r.

n.r. n.r.

n.r. n.r.

12 (12) 12 (12) 12 (12) 12 (6)

22* 3 8 14 1·5 12 5 4 3 1·5

7 to 3 (−4) 7 to 1 (−6) 4 to 0 (−4) n.r. 5 to 1 (−4) 8 to 2 (−6)*

0·082 0·029 < 0·001 n.r. < 0·001 n.r.

63 59 71 n.r. n.r. 82

n.r. – n.r. –

– 13 to 9 (−4)* 13 to 8 (−5)* 13 to 9 (−4)* 12 to 8 (−4)* 12 to 6 (−6)* 13 to 7 (−6) – 12 to 9 (−3) –

12 (12–14)

5 (1·5–22)

(–4 (−4 to −6))

71 (63–82)

31

13 to 8 (−4 (−3 to −6))

P

< 0·001 0·001 0·001 < 0·001 0·001 < 0·001 < 0·001

All values have been calculated to the nearest integer. *Mean values. †From baseline to follow-up; difference is shown in parentheses. ‡Outcome data reported in groups only. FI, faecal incontinence; CCIS, Cleveland Clinic Incontinence Score; n.r., outcome measured but not reported.

Results of studies of percutaneous tibial nerve stimulation PTNS studies included six case series31,33,43 – 46 (one study31 included a ‘control’ group for comparison) and one small randomized single-blind trial (PTNS versus TTNS versus sham) of patients from a single centre37 , comprising a total of 213 patients who received active treatment. All studies treated patients in 12 or 14 30min sessions, with four studies33,43,45,46 performing weekly treatments for 12 weeks, two studies37,44 performing twiceweekly treatments for 6 weeks and one study31 performing treatment every other day for 4 weeks. The median length of follow-up was 5 (range 1·5–22) months from start of treatment, indicating that two studies37,46 followed up patients only immediately after treatment had ended. Four studies31,33,43,44 then offered ‘top-up’ treatments to those who perceived treatment benefit (range 3–72 topups). Outcomes of patients receiving PTNS are summarized in Tables 5 and 6 (full version of Table 5 is supplied as Table S5, supporting information); further details of treatment protocols are given in Table S2 (supporting information).

Primary outcome Three studies33,37,44 reported a 50 per cent or greater reduction in the number of FI episodes immediately after treatment, in 63–82 per cent of patients. In the one study44 that reported on this outcome after 1 year, 59 per cent of patients still experienced treatment success (Table 5).  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

Secondary clinical outcomes Two studies33,37 reported rates of complete continence following treatment of 39 and 82 per cent. Statistically significant reductions in median weekly FI episodes over the three studies33,44,46 that reported this outcome were between 4 and 6 episodes per week. One study37 reported a mean reduction in weekly FI episodes of 6, but the significance of this was not reported (Table 5). The CCIS was used as an outcome measure in four studies33,43,44,46 , with two33,46 reporting a significant reduction in median CCIS and two43,44 a significant reduction in the mean score (all P ≤ 0·001) (Table 5). Quality of life Five studies31,37,43 – 45 reported changes in QoL after treatment, especially in the domains of depression, coping/behaviour, embarrassment and lifestyle, although different methods were used to collect the data. Other outcome measures are also shown in Table 6. Studies of transcutaneous tibial nerve stimulation TTNS studies included four non-randomized case series47 – 50 and two RCTs (one of TTNS versus sham36 and the other of PTNS versus TTNS versus sham37 ), comprising a total of 162 patients who received the active treatment. All studies used different treatment protocols, with a median treatment duration of 2·25 (range 1–3) months involving a median of 73 (12–180) treatments. Follow-up was for a median of 3·5 (1·5–15) months. www.bjs.co.uk

BJS 2014; 101: 457–468

Tibial nerve stimulation in faecal incontinence

Table 6

463

Other outcomes in trials of percutaneous tibial nerve stimulation

Reference

Outcome

Shafik et al.31

Rectometrogram Recurrence of symptoms

de la Portilla et al.43

FIQL score

VAS for QoL Anorectal manometry

SF-36

Govaert et al.44

FIQL score

Boyle et al.33

Time to defer defaecation Duration of effect Outcome versus physiological parameters

Findlay et al.45

Hotouras et al.46 George et al.37

HAD score ICIQ-B score FIQL score Time to defer defaecation St Mark’s incontinence score FIQL score SF-36 QoL score Anorectal manometry

Results No clear result shown Recurrence of symptoms in 30% of patients who originally had success, and with further treatment 75% had improved symptoms again FI no longer affected QoL 6 months after treatment in 12·5% of patients (reduction from 37·5% of patients to 25%) Significant improvement in FIQL score in depression, coping/behaviour and embarrassment domains, but not in lifestyle 6 months after treatment versus baseline (P < 0·004, P < 0·02, P < 0·005 and P = 0·086 respectively)* Significant improvement at 3 months and after 6 months without treatment (P = 0·002 and P = 0·001 respectively)* Significant improvement found in mean squeeze pressure between baseline and 6 months after treatment (P < 0·007), but not in resting pressures or rectal sensation SF-36 improved significantly in all domains apart from vitality at 1 year (P < 0·005)* FIQL score improved significantly in coping/behaviour and embarrassment domains after treatment, and in lifestyle and coping/behaviour domains at 1 year (P < 0·005)* Deferment of defaecation time improved in 65% patients from a median of 1 (range 0–15) to 5 (0–25) min (P < 0·001) Duration of effect data not collected or reported systematically No correlation found between outcome and physiological parameters Significant improvements seen in ICIQ-B score bowel control and QoL domains (P = 0·001 and P = 0·007 respectively), and in FIQL score lifestyle domain (P = 0·028)*. All other parameters showed non-significant improvements Improved significantly from 1 (range 0–30) min before treatment to 5 (0–60) min after treatment (P < 0·001) Improvement in continence score seen, but no comment on significance* Improvement in FIQL score, but no comment on significance* Improvement in SF-36 score, but no comment on significance* Significant improvement in mean(s.d.) peak squeeze pressure from baseline to post-treatment from 50(29) to 68(34) cmH2 O (P = 0·043)

*Mean values reported in study. FIQL, Faecal Incontinence Quality of Life Scale; FI, faecal incontinence; QoL, quality of life; VAS, visual analogue scale; SF-36, Short Form 36 (QualityMetric, Lincoln, Rhode Island, USA); HAD, Hospital Anxiety and Depression Scale; ICIQ-B, International Consultation on Incontinence Questionnaire – Bowel module.

Two studies47,49 offered top-up treatments to patients who had perceived benefit from treatment. Outcomes of patients receiving TTNS are summarized in Tables 7 and 8 (full version of Table 7 is supplied as Table S6, supporting information); further details of treatment protocols are given in Table S3 (supporting information).

Primary outcome Two studies36,37 reported on the primary outcome; immediately after treatment 0 per cent of patients in one study36 and 45 per cent in the other37 had a greater than 50 per cent reduction in the number of FI episodes (Table 7). Secondary clinical outcomes Complete continence following treatment was reported by one study37 as 45 per cent and by another36 as 0  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

per cent. A significant reduction in the median number of weekly FI episodes was reported in the large multicentre RCT36 , from 1·7 to 1 (P = 0·004). The mean reduction in weekly FI episodes in the only other study37 reporting this outcome was 2·3 (from 7·4 to 5·1), although this study did not report significance. The CCIS was used as an outcome measure in five studies36,47 – 50 ; the median reduction in CCIS ranged from 0·5 to 10 immediately after treatment and was reported as significant in three studies36,49,50 (Table 7).

Quality of life Five studies36,37,48 – 50 reported QoL data before and after TTNS treatment, although different outcome measures were used. All domains showed significant improvement in QoL immediately after treatment. This effect tended to www.bjs.co.uk

BJS 2014; 101: 457–468

464

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

Faecal incontinence episodes and Cleveland Clinic Incontinence Score outcomes in trials of transcutaneous tibial nerve stimulation

Table 7

Follow-up (months from start of treatment)

n*

Initial treatment period (months)‡

Queralto et al.47

10 (10)

1 (20)

Vitton et al.48** Vitton et al.49

12 (9) 24 (22)

3 (90) 3 (90)

Eleouet et al.50

32 (30)

1 (56)

Leroi et al.36 George et al.37

73 (68) 11 (n.r.)

3 (180) 1·5 (12)

1 4 3 3 15† 1 3 6 3 1·5

162

1–3 (12–180)

1·5–15

Reference

Summary (range)

FI episodes/week

CCIS ≥ 50% improvement in CCIS (% of patients)

Median (range)§

P

≥ 50% reduction (%)

–

–

–

– –

– –

– –

–

–

–

60 70 8 4 8 16

0 45

47¶ –

11 to 8 (−3) –

0–45

4–70

(–0·5 to −11·5)

1·7 to 1·0 (−0·7) 0·004 7·4 to 5·1 (–2·3)† n.r. (−0·7 to –2·3)

Median CCIS§

P

13 to 3 (−10) n.r. 13 to 1·5 (−11·5) n.r. 13·5 to 13 (−0·5) n.s. 14 to 12 (−2) 0·025 12 to 10 (−2) n.r. 14·5 to 11 (−3·5)† < 0·001

< 0·001 –

*Values in parentheses are numbers of women. †Data reported as mean values. ‡Number of initial treatments shown in parentheses. §From baseline to follow-up; difference is shown in parentheses. ¶More than 30 per cent improvement in Wexner score. #Therapy continued at home. **Patients with stable inflammatory bowel disease. ††Mean score improved. FI, faecal incontinence; CCIS, Cleveland Clinic Incontinence Score; n.r., outcome measured but not reported; n.s., not significant. Table 8

Other outcomes in trials of transcutaneous tibial nerve stimulation

Reference Queralto et al.47 Vitton et al.48

Vitton et al.49 Eleouet et al.50

Outcome Anorectal manometry VAS for symptoms VAS for QoL Time to defer defaecation VAS symptom improvement GIQLI VAS for subjective improvement FIQL score KESS score ADS score

Leroi et al.36

George et al.37

FIQL score Anorectal manometry

St Mark’s incontinence score

FIQL score SF-36 QoL score Anorectal manometry

Results No change from baseline to post-treatment 42% reported symptomatic improvement at 3 months Improved by 50% in same 42% of patients as above Improvement reported in 25% of patients 54% reported symptomatic improvement at 3 months Not predictive of treatment success. No other data presented Mean VAS: 63% perceived improvement at 1 month with mean improvement of 25% (±30%) Mean FIQL score improved significantly in all areas at 3 months but only coping/behaviour and general at 6 months Mean KESS score improved significantly at 3 months but not at 6 months Mean ADS anxiety score improved significantly at 3 months but not at 6 months Marked improvement in all four domains Significant improvement in maximum squeeze pressure after 3 months’ treatment (but not resting pressure, squeeze duration or rectal sensitivity) Improvement in incontinence score from mean(s.d.) 18·5(3·1) to 14·7(6·7) after 3 months; reduction in bowel movements from 2·6(1·4) to 2·4(1·3) per 24 h; increase in deferment to defaecation time from 1·9(0·9) to 2·2(2·4) min (no statistical analysis reported) Improvements in all domains of SF-36 (no statistical analysis reported) Improvement in peak squeeze pressure after 3 months’ treatment (but not resting pressure, rectal or anal sensitivity)

VAS, visual analogue scale; QoL, quality of life; GIQLI, Gastrointestinal Quality of Life Index; FIQL, Faecal Incontinence Quality of Life Scale; KESS, Knowles–Eccersley–Scott Symptom Score; ADS, Anxiety and Depression Scale; SF-36, Short Form 36 (QualityMetric, Lincoln, Rhode Island, USA).

reduce over time. Other outcome measures are also shown in Table 8.

Discussion

The two randomized trials36,37 were appraised using the Jadad score41 and both scored 4 of a possible 5 marks,  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

indicating good quality. These RCTs were also appraised using the Cochrane Collaboration tool42 for assessing risk of bias in randomized trials. The three-arm trial comparing PTNS versus TTNS versus sham had two areas of potential bias, in blinding and suitability of the sham procedure. This was purported to be a single (patient-only)-blind study. Although the practitioner giving the treatment www.bjs.co.uk

BJS 2014; 101: 457–468

Tibial nerve stimulation in faecal incontinence

465

was necessarily unblinded, final data collection was not, as should have been possible, performed by a blinded researcher. The paper acknowledges that the sham arm of the trial was suitable only for the TTNS arm, leading to difficulty in drawing conclusions between PTNS and sham results. Furthermore, patients in the ‘sham’ arm received a short burst (30 s) of treatment with TENS, followed by the machine being turned off; it was therefore possible that patients were unblinded to allocation, and this was not formally assessed. The small sample size, with only 30 participants in a three-arm study, although based on a prior sample size calculation, must also be noted. In the larger RCT36 comparing TTNS with sham stimulation, one area of potential bias, that of incomplete outcome data, was identified. Thirteen patients did not complete the treatment and were excluded from final analysis, with no explanation for drop-outs or indication of their original allocation. Formal intention-to-treat analysis was not performed. Sham stimulation with a placebo stimulator, resembling the active stimulator in appearance but with no delivery of any electrical stimulation, is also questionable. Although the paper states that patients were not aware that ‘sham’ stimulation was being tested, they may well have suspected this. Critical appraisal of the case series, using NICE Quality Assessment for Case Series, showed that all studies had a score of between 3 and 6 of 8. The median score for TTNS studies47 – 50 was 4·5 (range (3–6) and that for the PTNS studies 4 (3–5). Studies lost marks mostly because they were single-centre studies or patients were not recruited consecutively; results were often poorly reported with outcomes not stratified. The ongoing problem concerning lack of a standardized and universally accepted outcome measure also exists. This problem includes not only what outcome measure(s) is chosen, but also the way in which this measure is presented and interpreted. Most studies used the CCIS or bowel diary data (or both) as their primary outcome measure, and this is reasonable. The methods by which data have been analysed and presented, and the lack of inclusion of raw data, pose a problem. The presentation and interpretation of bowel diary data made comparison between studies difficult. Many of the included studies compared the group mean or median number of pretreatment and post-treatment FI episodes to assess for significant improvement. Weekly FI episodes, as a count, give an overdispersed Poisson distribution, that is with greater variability than expected. Attempting to define a clinically significant mean reduction in FI episodes per week in a population of patients with widely dispersed starting FI frequencies is very difficult. Therefore, although a significant result was demonstrated

in these studies, this does not necessarily correlate with a clinically significant result. Other studies chose to counter this problem by adopting a categorical measure of percentage reduction (the proportion of patients with a reduction in FI episodes per week of 50 per cent or more), which is probably a much more realistic indication of success, but unfortunately not universally reported. Although subject to criticism, this was chosen as the primary outcome for the present review, not least because it has been used most often to assess SNS, thus allowing comparisons to be drawn between the two treatment modalities. CCIS values have been compared by testing for statistical differences between the mean or median pretreatment and post-treatment scores. Analysis in this way gives no indication of individual patient experiences – the proportion that experienced improvement or deterioration, or the magnitude of individual improvement. As this outcome denotes a score, it also indicates nothing about how many are actually ‘better’ or ‘cured’. A better way to interpret these data may be to indicate the proportion of patients with a particular or significant improvement (for example, improved by 25, 50, 75 or 100 per cent), which some studies have done, but again this was not universal. Studies were generally hampered by a lack of common treatment indication, protocol denoting treatment timing and duration, and follow-up period. Although some studies reported patients who failed conservative treatment strategies, others did not. Similarly, in studies that required patients to have failed conservative treatment, those treatments were not discussed or standardized. No study analysed results separately based on patient sex, so comment on the success of treatment by sex is not possible. The same is true for incontinence type (urge or passive FI), and for the influence of presence or absence of anal sphincter defects. Treatment protocols for PTNS were fairly standardized. All used one of two sets of equipment discussed above; the number of treatments was fairly constant throughout studies and was performed once or twice weekly in all but one study31 . The presence, frequency and duration of top-up treatments, however, were different across all studies. No conclusions can therefore be drawn regarding which treatment protocol is most efficacious. For TTNS therapy, although similar equipment was used in all studies, each treatment protocol differed significantly in terms of the number of treatments, frequency and duration, as well as top-up treatment protocols. Patients receiving TTNS had up to 180 treatments, which is obviously a factor of ten greater than most PTNS treatment protocols included in this systematic review. One potential advantage of TTNS over PTNS is the simplicity of treatment administration, resulting in the possibility of

 2014 BJS Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

BJS 2014; 101: 457–468

466

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

patient-administered home treatments; indeed, one study49 of TTNS discussed here involved patient-administered home treatments. This impacts on the potential patient acceptance of the treatment along with the associated costs, which again have to be borne in mind when comparing the two procedures. No study evaluated costs or costeffectiveness. The finding that no significant benefit of TTNS over sham stimulation was found in the study36 that involved 180 treatment administrations is particularly interesting. In summary, non-randomized observational studies of PTNS and TTNS showed improvements in most outcome measures after treatment compared with baseline. A small three-arm RCT37 of PTNS versus TTNS versus sham showed effect for both treatments over sham, with PTNS appearing to be superior. A much larger RCT36 of TTNS versus sham, however, refuted the finding for TTNS. A large UK-based multicentre RCT of PTNS versus sham is almost fully recruited51 and will resolve the question for PTNS. Further well conducted RCTs comparing PTNS and TTNS with sham or other proven therapies, using standardized protocols and outcome measures, are required to provide high-quality clinical evidence of a definitive treatment algorithm for FI. Based on current high-quality evidence, however, it seems that TTNS is not superior to sham electrical stimulation and there is insufficient evidence to draw conclusions about the effectiveness of PTNS compared with sham. Disclosure

The authors declare no conflict of interest. References 1 Nelson RL. Epidemiology of fecal incontinence. Gastroenterology 2004; 126(Suppl 1): S3–S7. 2 Perry S, Shaw C, McGrother C, Matthews RJ, Assassa RP, Dallosso H et al. Prevalence of faecal incontinence in adults aged 40 years or more living in the community. Gut 2002; 50: 480–484. 3 Edwards NI, Jones D. The prevalence of faecal incontinence in older people living at home. Age Ageing 2001; 30: 503–507. 4 Johanson JF, Lafferty J. Epidemiology of fecal incontinence: the silent affliction. Am J Gastroenterol 1996; 91: 33–36. 5 Madoff RD, Parker SC, Varma MG, Lowry AC. Faecal incontinence in adults. Lancet 2004; 364: 621–632. 6 Bharucha AE, Zinsmeister AR, Locke GR, Schleck C, McKeon K, Melton LJ. Symptoms and quality of life in community women with fecal incontinence. Clin Gastroenterol Hepatol 2006; 4: 1004–1009. 7 Kamm MA. Faecal incontinence. BMJ 1998; 316: 528–532.  2014 BJS Society Ltd Published by John Wiley & Sons Ltd

8 Collings S, Norton C. Women’s experiences of faecal incontinence: a study. Br J Community Nurs 2004; 9: 520–523. 9 Damon H, Guye O, Seigneurin A, Long F, Sonko A, Faucheron JL et al. Prevalence of anal incontinence in adults and impact on quality-of-life. Gastroenterol Clin Biol 2006; 30: 37–43. 10 Cotterill N, Norton C, Avery KN, Abrams P, Donovan JL. A patient-centered approach to developing a comprehensive symptom and quality of life assessment of anal incontinence. Dis Colon Rectum 2008; 51: 82–87. 11 Finne-Soveri H, Sorbye LW, Jonsson PV, Carpenter GI, Bernabei R. Increased work-load associated with faecal incontinence among home care patients in 11 European countries. Eur J Public Health 2008; 18: 323–328. 12 Miner PB Jr. Economic and personal impact of fecal and urinary incontinence. Gastroenterology 2004; 126(Suppl 1): S8–S13. 13 Norton C, Cody JD. Biofeedback and/or sphincter exercises for the treatment of faecal incontinence in adults. Cochrane Database Syst Rev 2012; (7)CD002111. 14 Sun WM, Read NW, Verlinden M. Effects of loperamide oxide on gastrointestinal transit time and anorectal function in patients with chronic diarrhoea and faecal incontinence. Scand J Gastroenterol 1997; 32: 34–38. 15 Malouf AJ, Norton CS, Engel AF, Nicholls RJ, Kamm MA. Long-term results of overlapping anterior anal-sphincter repair for obstetric trauma. Lancet 2000; 355: 260–265. 16 Tillin T, Gannon K, Feldman RA, Williams NS. Third-party prospective evaluation of patient outcomes after dynamic graciloplasty. Br J Surg 2006; 93: 1402–1410. 17 National Institute for Health and Care Excellence (NICE). Faecal Incontinence: the Management of Faecal Incontinence in Adults. NICE Clinical Guidelines CG49. http://www.nice.org.uk/CG49 [accessed 24 November 2013]. 18 Brown SR, Wadhawan H, Nelson RL. Surgery for faecal incontinence in adults. Cochrane Database Syst Rev 2013; (7)CD001757. 19 Mowatt G, Glazener C, Jarrett M. Sacral nerve stimulation for faecal incontinence and constipation in adults. Cochrane Database Syst Rev 2007; (3)CD004464. 20 Leroi AM, Parc Y, Lehur PA, Mion F, Barth X, Rullier E et al. Efficacy of sacral nerve stimulation for fecal incontinence: results of a multicenter double-blind crossover study. Ann Surg 2005; 242: 662–669. 21 Tjandra JJ, Chan MKY, Yeh CH, Murray-Green C. Sacral nerve stimulation is more effective than optimal medical therapy for severe fecal incontinence: a randomized, controlled study. Dis Colon Rectum 2008; 51: 494–502. 22 Jarrett ME, Mowatt G, Glazener CM, Fraser C, Nicholls RJ, Grant AM et al. Systematic review of sacral nerve stimulation for faecal incontinence and constipation. Br J Surg 2004; 91: 1559–1569. ¨ 23 Matzel KE, Kamm MA, Stosser M, Baeten CG, Christiansen J, Madoff R et al. Sacral spinal nerve stimulation for faecal

www.bjs.co.uk

BJS 2014; 101: 457–468

Tibial nerve stimulation in faecal incontinence

24

25

26

27

28

29

30 31

32 33

34

35

36

37

38

incontinence: multicentre study. Lancet 2004; 363: 1270–1206. Wexner SD, Coller JA, Devroede G, Hull T, McCallum R, Chan M et al. Sacral nerve stimulation for fecal incontinence: results of a 120-patient prospective multicenter study. Ann Surg 2010; 251: 441–449. Thin NN, Horrocks EJ, Hotouras A, Palit S, Thaha MA, Chan CL et al. A systematic review of the clinical effectiveness of neuromodulation in the treatment of faecal incontinence. Br J Surg 2013; 100: 1430–1447. Dudding TC, Meng Lee E, Faiz O, Par´es D, Vaizey CJ, McGuire A et al. Economic evaluation of sacral nerve stimulation for faecal incontinence. Br J Surg 2008; 95: 1155–1163. Gladman MA, Knowles CH. Surgical treatment of patients with constipation and fecal incontinence. Gastroenterol Clin North Am 2008; 37: 605–625. Sprange K, Clift M, Burke M, Whitehead SR, Hutton J. Evidence Review: Sacral Nerve Stimulation for Faecal Incontinence. Healthcare Innovation and Technology Evaluation Centre (HITEC): Derby, 2009. McGuire EJ, Zhang SC, Horwinski ER, Lytton B. Treatment of motor and sensory detrusor instability by electrical stimulation. J Urol 1983; 129: 78–79. Stoller ML. Afferent nerve stimulation for pelvic floor dysfunction. Eur Urol 1999; 35: 16. Shafik A, Ahmed I, El-Sibai O, Mostafa RM. Percutaneous peripheral neuromodulation in the treatment of fecal incontinence. Eur Surg Res 2003; 35: 103–107. Allison M. Percutaneous tibial nerve stimulation for patients with faecal incontinence. Nurs Stand 2011; 25: 44–48. Boyle DJ, Prosser K, Allison ME, Williams NS, Chan CL. Percutaneous tibial nerve stimulation for the treatment of urge fecal incontinence. Dis Colon Rectum 2010; 53: 432–437. Findlay JM, Maxwell-Armstrong C. Posterior tibial nerve stimulation and faecal incontinence: a review. Int J Colorectal Dis 2011; 26: 265–273. Thomas GP, Dudding TC, Rahbour G, Nicholls RJ, Vaizey CJ. A review of posterior tibial nerve stimulation for faecal incontinence. Colorectal Dis 2013; 15: 519–526. Leroi AM, Siproudhis L, Etienney I, Damon H, Zerbib F, Amarenco G et al. Transcutaneous electrical tibial nerve stimulation in the treatment of fecal incontinence: a randomized trial (CONSORT 1a). Am J Gastroenterol 2012; 107: 1888–1896. George AT, Kalmar K, Sala S, Kopanakis K, Panarese A, Dudding TC et al. Randomized controlled trial of percutaneous versus transcutaneous posterior tibial nerve stimulation in faecal incontinence. Br J Surg 2013; 100: 330–338. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 2009; 339: b2700.

 2014 BJS Society Ltd Published by John Wiley & Sons Ltd

467

39 Jorge JM, Wexner SD. Etiology and management of fecal incontinence. Dis Colon Rectum 1993; 36: 77–97. 40 National Institute for Health and Care Excellence. Appendix 4: Quality Assessment for Case Series. http://www.nice.org.uk/ nicemedia/pdf/Appendix_04_qualityofcase_series_form_ preop.pdf [accessed 24 November 2013]. 41 Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996; 17: 1–12. 42 Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD et al.; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011; 343: d5928. 43 de la Portilla F, Rada R, Vega J, Gonzalez CA, Cisneros N, Maldonado VH. Evaluation of the use of posterior tibial nerve stimulation for the treatment of fecal incontinence: preliminary results of a prospective study. Dis Colon Rectum 2009; 52: 1427–1433. 44 Govaert B, Pares D, Delgado-Aros S, La Torre F, Van Gemert WG, Baeten CG. A prospective multicentre study to investigate percutaneous tibial nerve stimulation for the treatment of faecal incontinence. Colorectal Dis 2010; 12: 1236–1241. 45 Findlay JM, Yeung JM, Robinson R, Greaves H, MaxwellArmstrong C. Peripheral neuromodulation via posterior tibial nerve stimulation – a potential treatment for faecal incontinence? Ann R Coll Surg Engl 2010; 92: 385–390. 46 Hotouras A, Thaha MA, Allison ME, Currie A, Scott SM, Chan CL. Percutaneous tibial nerve stimulation (PTNS) in females with faecal incontinence: the impact of sphincter morphology and rectal sensation on the clinical outcome. Int J Colorectal Dis 2012; 27: 927–930. 47 Queralto M, Portier G, Cabarrot PH, Bonnaud G, Chotard JP, Nadrigny M et al. Preliminary results of peripheral transcutaneous neuromodulation in the treatment of idiopathic fecal incontinence. Int J Colorectal Dis 2006; 21: 670–672. 48 Vitton V, Damon H, Roman S, Nancey S, Flouri´e B, Mion F. Transcutaneous posterior tibial nerve stimulation for fecal incontinence in inflammatory bowel disease patients: a therapeutic option? Inflamm Bowel Dis 2009; 15: 402–405. 49 Vitton V, Damon H, Roman S, Mion F. Transcutaneous electrical posterior tibial nerve stimulation for faecal incontinence: effects on symptoms and quality of life. Int J Colorectal Dis 2010; 25: 1017–1020. 50 Eleouet M, Siproudhis L, Guillou N, Le Couedic J, Bouguen G, Bretagne JF. Chronic posterior tibial nerve transcutaneous electrical nerve stimulation (TENS) to treat fecal incontinence (FI). Int J Colorectal Dis 2010; 25: 1127–1132. 51 NHS National Institute for Health Research. HTA – 09/104/16: CONtrol of Faecal Incontinence using Distal NeuromodulaTion (CONFIDeNT). http://www.nets.nihr. ac.uk/projects/hta/0910416 [accessed 29 November 2013].

www.bjs.co.uk

BJS 2014; 101: 457–468

468

E. J. Horrocks, N. Thin, M. A. Thaha, S. J. C. Taylor, C. Norton and C. H. Knowles

Supporting information

Additional supporting information may be found in the online version of this article: Table S1 Characteristics of randomized clinical trials (Word document) Table S2 Characteristics of percutaneous tibial nerve stimulation studies (Word document) Table S3 Characteristics of transcutaneous tibial nerve stimulation studies (Word document) Table S4 Full version of Table 3 (Word document) Table S5 Full version of Table 5 (Word document) Table S6 Full version of Table 7 (Word document)

Snapshot quiz

Snapshot quiz 14/5 Question: What is this lesion and what is the management?

The answer to the above question is found on p. 494 of this issue of BJS. Brown C, Chamary V: Department of General Surgery, Royal Gwent Hospital, Cardiff Road, Newport NP20 2UB, UK (e-mail: [email protected])

Snapshots in Surgery: to view submission guidelines, submit your snapshot and view the archive, please visit www.bjs.co.uk

 2014 BJS Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

BJS 2014; 101: 457–468