Supplemental materials

eTable 1. MOOSE Checklist

eTable 2. General Search Strategies for PubMed and Embase.

eTable 3. List of Included and Excluded Studies.

eTable 4. The Critical Appraisal of Included Studies Using Newcastle-Ottawa Scale (NOS).

eTable 5. Grading Evidence Based on Egger’s P value, Sample Size and Heterogeneity.

eTable 6. Rate of Anal Fistula Recurrence by Study-level Factors.

eTable 7. Definition of Anal Fistula Recurrence from Each Original Study.

eTable 8. Sensitivity Analysis for Significant Factors and Class of Evidence.

eTable 9. Sensitivity Analysis for Non-significant Factors and Class of Evidence.

eTable 10. Risk Factors of Anal Fistula Recurrence in Patients with Mucosal Advancement Flap Surgery.

eTable 11. Sensitivity Analysis for Outcomes of Anal Fistula Recurrence Associated with Patient-related Factors, Fistula and Surgery-related factors.

eTable 12. Sensitivity Analysis for Risk Factors of Anal Fistula Recurrence in Patients with Mucosal Advancement Flap Surgery and Class of Evidence.

eFigure 1. Forest Plot for Pooled Rate of Anal Fistula Recurrence.

eTable 1. MOOSE Checklist

eTable 2- MOOSE Checklist

Item No RecommendationReportedon Page

NoReporting of background should include

1 Problem definition 6-7

2 Hypothesis statement 7

3 Description of study outcome(s) 7

4 Type of exposure or intervention used 7

5 Type of study designs used 7

6 Study population 7

Reporting of search strategy should include

7 Qualifications of searchers (eg, librarians and investigators) 1

8 Search strategy, including time period included in the synthesis and key words 8

9 Effort to include all available studies, including contact with authors 8

10 Databases and registries searched 7-8

11 Search software used, name and version, including special features used (eg, explosion) eTable 2

12 Use of hand searching (eg, reference lists of obtained articles) 8

13 List of citations located and those excluded, including justification eTable 3

14 Method of addressing articles published in languages other than English 8

15 Method of handling abstracts and unpublished studies 8

16 Description of any contact with authors 8

Reporting of methods should include

17 Description of relevance or appropriateness of studies assembled for assessing thehypothesis to be tested 9

18 Rationale for the selection and coding of data (eg, sound clinical principles or convenience) 9

19 Documentation of how data were classified and coded (eg, multiple raters, blinding andinterrater reliability) 10

20 Assessment of confounding (eg, comparability of cases and controls in studies whereappropriate) x

1

21 Assessment of study quality, including blinding of quality assessors, stratification or regression on possible predictors of study results 10

22 Assessment of heterogeneity 11-12

23

Description of statistical methods (eg, complete description of fixed or random effects models, justification of whether the chosen models account for predictors of study results, dose-response models, or cumulative meta-analysis) in sufficient detail to be replicated

11-12

24 Provision of appropriate tables and graphics x

Reporting of results should include

25 Graphic summarizing individual study estimates and overall estimate Figure 226 Table giving descriptive information for each study included Table 1

27 Results of sensitivity testing (eg, subgroup analysis) eTable 8,9,11,12

28 Indication of statistical uncertainty of findings x

Item No RecommendationReportedon Page

NoReporting of discussion should include

29 Quantitative assessment of bias (eg, publication bias) eTable 8, eTable 9

30 Justification for exclusion (eg, exclusion of non-English language citations) x

31 Assessment of quality of included studies 17-19

Reporting of conclusions should include

32 Consideration of alternative explanations for observed results 17-18

33 Generalization of the conclusions (ie, appropriate for the data presented and within the domain of the literature review) 22

34 Guidelines for future research 21-22

eTable 2. General Search Strategies for PubMed and Embase.

PubMed Search Strategy 1. "Rectal Fistula"[Mesh]2. ‘anal fistul*’ OR ‘anus fistul*’ OR ‘fistula-in-ano’ OR ‘anorectal fistul*’ OR ‘perianal fistul*’[Title/Abstract]3. 1 OR 24. "Recurrence"[Mesh]5. "Treatment Outcome"[Mesh]6. "Treatment Failure"[Mesh]7. recur* or relaps* or remission* [Title/Abstract]8. 4-7/OR9. 3 AND 8 Embase Search Strategy 1. 'rectum fistula'/exp2. (‘anal fistul*’ OR ‘anus fistul*’ OR fistula-in-ano OR ‘anorectal fistul*’ OR ‘perianal fistul*’):ab,ti3. 1 OR 24. 'recurrent disease'/exp5. 'treatment outcome'/exp6. 'treatment failure'/exp7. (recur* or relaps* or remission*):ab,ti8. 4-7/OR9. 3 AND 8

eTable 3. List of Included and Excluded Studies.

Included and excluded reasons No. studies References.

Included studies 20 [1-20]1-20

Studies including IBD or rectovaginal fistula patients

5 [21-25]21-25

Letters, comments or case reports 9 [26-34]26-34

Studies without outcome data 13 [35-47]35-47

Studies with insufficient data 3 [48-50] 48-50

Abbreviations: IBD, inflammatory bowel disease.

REFERENCES

1. Placer Galán C, Lopes C, Múgica J, et al. Patterns of recurrence/persistence of criptoglandular anal fistula after the LIFT procedure. Long-term observacional study. Cir Esp 2017;95:385-390.

2. Emile S, Elfeki H, Thabet W, et al. Predictive factors for recurrence of high transsphincteric anal fistula after placement of seton. J. Surg. Res. 2017;213:261-268.

3. Boenicke L, Karsten E, Zirngibl H, et al. Advancement Flap for Treatment of Complex Cryptoglandular Anal Fistula: Prediction of Therapy Success or Failure Using Anamnestic and Clinical Parameters. World J Surg 2017;41:2395-2400.

4. Visscher AP, Schuur D, Slooff RAE, et al. Predictive factors for recurrence of cryptoglandular fistulae characterized by preoperative three-dimensional endoanal ultrasound. Colorectal Disease 2016;18:503-509.

5. Parthasarathi R, Gomes RM, Rajapandian S, et al. Ligation of the intersphincteric fistula tract for the treatment of fistula-in-ano: experience of a tertiary care centre in South India. Colorectal Dis 2016;18:496-502.

6. Mijnsbrugge GV, Deen-Molenaar C, Ho D, et al. LIFT procedure for peri-anal fistulas: Risk factors for failure. Colorectal Disease 2016;18:17.

7. Li J, Yang W, Huang Z, et al. [Clinical characteristics and risk factors for recurrence of anal fistula patients]. Zhonghua Wei Chang Wai Ke Za Zhi 2016;19:1370-1374.

8. Schulze B, Ho YH. Management of complex anorectal fistulas with seton drainage plus partial fistulotomy and subsequent ligation of intersphincteric fistula tract (LIFT). Tech Coloproctol 2015;19:89-95.

9. Ding JH, Bi LX, Zhao K, et al. Impact of three-dimensional endoanal ultrasound on the outcome of anal fistula surgery: a prospective cohort study. Colorectal Dis 2015;17:1104-12.

10. van Onkelen RS, Gosselink MP, Thijsse S, et al. Predictors of outcome after transanal advancement flap repair for high transsphincteric fistulas. Dis Colon Rectum 2014;57:1007-11.

11. Schwandner O. Obesity is a negative predictor of success after surgery for complex anal fistula. BMC Gastroenterol 2011;11:61.

12. Devaraj B, Khabassi S, Cosman BC. Recent smoking is a risk factor for anal abscess and fistula. Dis Colon Rectum 2011;54:681-5.

13. Abbas MA, Jackson CH, Haigh PI. Predictors of outcome for anal fistula surgery. Arch Surg 2011;146:1011-6.

14. Yano T, Asano M, Matsuda Y, et al. Prognostic factors for recurrence following the initial drainage of an anorectal abscess. Int J Colorectal Dis 2010;25:1495-8.

15. Jordán J, Roig JV, García-Armengol J, et al. Risk factors for recurrence and incontinence after anal fistula surgery. Colorectal Disease 2010;12:254-260.

16. Hamadani A, Haigh P, Liu I, et al. Who is at risk for developing chronic anal fistula or recurrent anal sepsis after initial perianal abscess? Dis. Colon Rectum 2009;52:217-21.

17. van Koperen PJ, Wind J, Bemelman WA, et al. Long-term functional outcome and risk factors for recurrence after surgical treatment for low and high perianal fistulas of cryptoglandular origin. Dis Colon Rectum 2008;51:1475-81.

18. Poon C, Chi-Ming P, Ng D, et al. Recurrence pattern of fistula-in-ano in a Chinese population. J Gastrointestin Liver Dis 2008;17:53-7.

19. Zimmerman DD, Delemarre JB, Gosselink MP, et al. Smoking affects the outcome of transanal mucosal advancement flap repair of trans-sphincteric fistulas. Br J Surg 2003;90:351-4.

20. Garcia-Aguilar J, Belmonte C, Wong WD, et al. Anal fistula surgery. Factors associated with recurrence and incontinence. Dis Colon Rectum 1996;39:723-9.

21. Ajayi AO, Chandrasekar T, Hammed AH. Crohn's disease presenting as a recurrent perianal fistula: a case report. Niger J Clin Pract 2010;13:473-6.

22. Angelberger S, Reinisch W, Dejaco C, et al. NOD2/CARD15 gene variants are linked to failure of antibiotic treatment in perianal fistulating Crohn's disease. Am J Gastroenterol 2008;103:1197-202.

23. Ardizzone S, Maconi G, Colombo E, et al. Perianal fistulae following infliximab treatment: clinical and endosonographic outcome. Inflamm Bowel Dis 2004;10:91-6.

24. Chung W, Ko D, Sun C, et al. Outcomes of anal fistula surgery in patients with inflammatory bowel disease. Am J Surg 2010;199:609-13.

25. van Koperen PJ, Safiruddin F, Bemelman WA, et al. Outcome of surgical treatment for fistula in ano in Crohn's disease. Br J Surg 2009;96:675-9.

26. Dunn K, Pasternak B, Kelsen JR, et al. Mevalonate kinase deficiency presenting as recurrent rectal abscesses and perianal fistulae. Ann Allergy Asthma Immunol 2018;120:214-215.

27. Ferreira Cardoso M, Carneiro C, Carvalho Lourenco L, et al. Actinomycosis Causing Recurrent Perianal Fistulae. ACG Case Rep J 2017;4:e82.

28. Fry RD, Birnbaum EH, Lacey DL. Actinomyces as a cause of recurrent perianal fistula in the immunocompromised patient. Surgery 1992;111:591-4.

29. Hasan A, Evgenikos N, Daniel T, et al. Filshie clip migration with recurrent perianal sepsis and low fistula in ano formation. Bjog 2005;112:1581.

30. Johnson KN, Young-Fadok TM, Carpentieri D, et al. Case report: misdiagnosis of tailgut cyst presenting as recurrent perianal fistula with pelvic abscess. J Pediatr Surg 2013;48:e33-6.

31. Mayoral JL, Rubal BJ. Unusual sonographic finding in a patient with late recurrence of a perianal fistula. J Clin Ultrasound 2002;30:557-61.

32. Sauer J, Wolf HK, Junginger T. [Adenocarcinoma in a tail-gut cyst: a rare cause of recurrent perianal fistula]. Chirurg 2000;71:712-6.

33. Zbar A. Long-term outcome following loose-seton technique for external sphincter preservation in complex anal fistula (Br J Surg 2004; 91: 476-480). Br J Surg 2004;91:1073.

34. Zimmerman DD, Mitalas LE, Schouten WR. Long-term functional outcome and risk factors for recurrence after surgical treatment for low and high perianal fistulas of cryptoglandular origin. Dis Colon Rectum 2009;52:1196-7; author reply 1197.

35. Abramowitz L, Soudan D, Souffran M, et al. The outcome of fistulotomy for anal fistula at 1 year: a prospective multicentre French study. Colorectal Dis 2016;18:279-85.

36. Balogh G. ["Loop"-drainage in the management of recurrent extra-sphincteric perianal fistulae]. Orv Hetil 1994;135:2705-8.

37. Balogh G. Tube loop (seton) drainage treatment of recurrent extrasphincteric perianal fistulae. Am J Surg 1999;177:147-9.

38. Buchanan GN, Halligan S, Bartram CI, et al. Clinical examination, endosonography, and MR imaging in preoperative assessment of fistula in ano: comparison with outcome-based reference standard. Radiology 2004;233:674-81.

39. Christiansen J, Ronholt C. Treatment of recurrent high anal fistula by total excision and primary sphincter reconstruction. Int J Colorectal Dis 1995;10:207-9.

40. Fisher OM, Raptis DA, Vetter D, et al. An outcome and cost analysis of anal fistula plug insertion vs endorectal advancement flap for complex anal fistulae. Colorectal Dis 2015;17:619-26.

41. Gingold BS. Reducing the recurrence risk of fistula in ano. Surg Gynecol Obstet 1983;156:661-2.42. Milone M, Pesce G, Leongito M, et al. [Role of endoanal ultrasonography in reducing anal fistula

recurrence]. Chir Ital 2009;61:461-5.43. Ozturk E. Treatment of recurrent anal fistula using an autologous cartilage plug: a pilot study. Tech

Coloproctol 2015;19:301-7.44. Safar B, Jobanputra S, Sands D, et al. Anal fistula plug: initial experience and outcomes. Dis Colon

Rectum 2009;52:248-52.45. Spencer JA, Chapple K, Wilson D, et al. Outcome after surgery for perianal fistula: predictive value of

MR imaging. AJR Am J Roentgenol 1998;171:403-6.46. Tan KK, Kaur G, Byrne CM, et al. Long-term outcome of the anal fistula plug for anal fistula of

cryptoglandular origin. Colorectal Dis 2013;15:1510-4.47. Waniczek D, Adamczyk T, Arendt J, et al. Direct MRI fistulography with hydrogen peroxide in patients

with recurrent perianal fistulas: a new proposal of extended diagnostics. Med Sci Monit 2015;21:439-45.48. Gottgens KW, Janssen PT, Heemskerk J, et al. Long-term outcome of low perianal fistulas treated by

fistulotomy: a multicenter study. Int J Colorectal Dis 2015;30:213-9.49. Tao Y, Yang G, Qiu J, et al. [Association of serum miRNA6086 expression with postoperative anal fistula

recurrence]. Zhonghua Wei Chang Wai Ke Za Zhi 2017;20:1409-1413.50. van der Hagen SJ, Baeten CG, Soeters PB, et al. Long-term outcome following mucosal advancement

flap for high perianal fistulas and fistulotomy for low perianal fistulas: recurrent perianal fistulas: failure of treatment or recurrent patient disease? Int J Colorectal Dis 2006;21:784-90.

eTable 4. The Critical Appraisal of Included Studies Using Newcastle-Ottawa Scale (NOS)

Score by NOS CategoryNo. Study, by First

AuthorSelection Comparability Outcome/

ExposureOverall Score

1 Placer Galán et al 2017

2 1 3 6

2 Emile et al 2017 2 2 3 73 Boenicke et al 2017 3 2 3 84 Parthasarathi et al

20163 1 3 7

5 Visscher et al 2016 3 1 3 76 Mijnsbrugge et al

20163 0 3 6

7 Li et al 2016 3 0 1 48 Ding et al 2015 3 1 3 79 Schulze et al 2015 3 0 3 610 Zimmerman et al

20033 1 2 6

11 van Onkelen et al 2014

3 2 3 8

12 Devaraj et al 2011 3 1 2 613 Abbas et al 2011 3 1 2 614 Schwandner et al

20113 1 3 7

15 Yano et al 2010 3 1 1 516 Jordán et al 2010 3 2 3 817 Hamadani et al

20093 2 2 7

18 Poon et al 2008 3 1 2 6

19 van Koperen et al 2008

3 2 3 8

20 Garcia-Aguilar et al 1996

2 1 2 5

eTable 5. Grading Evidence Based on Egger’s P value, Sample Size and Heterogeneity.

In order to increase the comprehensibility of the findings, the effect of risk factors was graded by Egger’s P value, total sample size and heterogeneity using a set of modified criteria.

Evidence DefinitionClass 1 (High-quality) evidence

Defined as three conditions of Egger’s P value of the random effects model >0.1, a total population >1000, and lower between-study heterogeneity I2<50% are met simultaneously.

Class II (Moderate-quality) evidence

Defined as two of the three conditions of Egger’s P value of the random effects

model >0.1, a pooled population >1000, and higher between-study heterogeneity I2

≥50% are met.Class III (Moderate-quality) evidence

Defined as one of the three conditions of Egger’s P value of the random effects model >0.1, a pooled population <1000, and lower between-study heterogeneity I2<50% are met.

Class IV (Low-quality) evidence

Defined as none of the three conditions of Egger’s P value of the random effects model >0.1, a pooled population <1000, and higher between-study heterogeneity I2≥50% are met.

eTable 6. Rate of Anal Fistula Recurrence by Study-level Factors.

Variable Rate, % 95% CI Degree ofheterogeneity

(I2 statistics; %)

P value No. of included Studies

P for interaction

Total　 0.19 0.15-0.23 96.8 <0.001 20 NASample size <0.001

＜100 0.28 0.11-0.46 91.1 <0.001 4100-500 0.20 0.14-0.27 95.5 <0.001 13

>500 0.05 0.02-0.09 95.1 <0.001 3Study region <0.001

Europe 0.28 0.17-0.39 95.4 <0.001 8USA 0.15 0.08-0.21 96.0 <0.001 5Asia 0.13 0.05-0.22 96.8 <0.001 5

Others 0.11 0.07-0.14 0.0 0.697 2Follow-up period (Median/mean)

<0.001

<12 months 0.37 -0.01-0.75 96.3 <0.001 212-24 months 0.17 0.09-0.25 95.3 <0.001 7>12 months 0.18 0.12-0.23 94.5 <0.001 9

Study design <0.001Prospective 0.18 0.13-0.24 94.7 <0.001 12

Retrospective 0.20 0.13-0.27 97.9 <0.001 8Institution involved 0.371

Single 0.21 0.15-0.26 97.3 <0.001 16Multiple 0.19 0.15-0.23 91.6 <0.001 4

Study year <0.001

~2000 0.05 0.03-0.07 - - 12000-2010 0.23 0.12-0.34 96.7 <0.001 6

2010~ 0.18 0.13-0.24 96.7 <0.001 13NOS score <0.001

Low (≤6) 0.18 0.13-0.24 96.7 <0.001 10High (＞6) 0.19 0.15-0.23 95.6 <0.001 10

Abbreviations: CI, confidence interval; NA, not available.

eTable 7. Definition of Anal Fistula Recurrence from Each Original Study.

Study Outcome DefinitionPlacer Galán et al., 2017 NAEmile et al., 2017 Recurrence was defined as the clinical persistence or reappearance of the fistulous track with or

without associated discharge after the removal of seton during follow-up.Boenicke et al., 2017 Fistula recurrence was defined as a fistula or an abscess occurring after initial healing.Parthasarathi et al., 2016 Any wound that had not healed at 6 weeks or any wound that healed and then presented with

purulent discharge, or any wound that had not healed at 6 weeks, was assessed by endoanal ultrasound and if a tract was found was defined as a recurrent fistula.

Visscher et al., 2016 There were symptoms of recurrence or any additional fistula surgery had been performed in other hospitals.

Mijnsbrugge et al., 2016 NALi et al., 2016 NADing et al., 2015 It was defined that the previously healed wound developed erythema, swelling, fluctuation, external

opening or fistula tract; (2) pain with tenderness and/or a mass on palpation around a previously healed fistula; and (3) a longstanding wound did not heal more than 3 months after surgery with the fistula tract or internal opening detected by EUA.

Schulze et al., 2015 Recurrences were classified as Type 1, a residual sinus tract from the external opening, Type 2, a

downstaged tract from transsphincteric to intersphincteric fistula, Type 3, a complete failure with the recurrent fistula tract extending from internal to external opening.

Zimmerman et al., 2015 Recurrence was defined that recurrent internal or external opening was observed 6 months later.van Onkelen et al., 2014 Healing of the fistula was defined as complete wound healing and closure of all external openings

in combination with absence of symptoms.Devaraj et al., 2011 NAAbbas et al., 2011 Operative failure or recurrence was defined as persistence or recurrence of symptoms within 6

months of intervention.Schwandner et al., 2011 Success was defined as closure of both internal and external openings, absence of drainage

without further intervention, and absence of abscess formation.Yano et al., 2010 Recurrence was defined as the recurrence of an abscess or the formation of an anal fistula.Jordán et al., 2010 NAHamadani et al., 2009 Recurrence of disease was defined as the development of recurrent perianal sepsis or chronic anal

fistulaPoon et al., 2008 Recurrence of disease was defined complete wound healing failed with persistent external opening

found 8 weeks after surgery, or 2) new external opening was found during follow up.van Koperen et al., 2008 The fistula was considered closed if the external opening was closed and no discharge or pain was

experienced. Otherwise, the fistulas were considered persistent or recurrent.Garcia-Aguilar et al., 1996

NA

eTable 8. Sensitivity Analysis for Significant Factors and Class of Evidence.

Significant factors Postoperative Recurrence RR (95% CI)

T-F adjusted RR (95% CI)

Filled studies

Class of Evidence

Prior anal surgeryNo Ref.

Yes 1.52 (1.04 to 2.23) 1.20 (0.82 to 1.76) 3 II

Surgical procedureFistulotomy Ref.

Seton placement 2.97 (1.10 to 8.06) 2.97 (1.10 to 8.06) 0 II

Fistula classificationLow transsphincteric Ref.High transsphincteric 4.77 (3.83 to 5.95) 4.17 (3.04 to 5.72) 3 I

Internal opening identifiedYes Ref.No 8.54 (5.29 to 13.80) 8.54(5.29 to 13.80) 0 I

Horseshoe extensionsNo Ref.Yes 1.92 (1.43 to 2.59) 1.71 (1.19 to 2.46) 2 I

No. of fistula tractsSingle Ref.

Multiple 4.77 (1.46 to 15.51) 4.77(1.46 to 15.51) 0 IIAbbreviations: NA, not available; Ref., Reference group; RR, relative risk; T-F, trim and filled method.

eTable 9. Sensitivity Analysis for Non-significant Factors and Class of Evidence.

Non-significant patient-related factors

Postoperative Recurrence RR (95% CI)

T-F adjusted RR (95% CI)

Filled studies

Class of Evidence

GenderFemale Ref.Male 1.00 (0.80 to 1.25) 0.93 (0.76 to 1.13) 5 I　 　Age (years)＞ 40 or 45 Ref.≤ 40 or 45 1.27 (0.99 to 1.62) 1.27 (0.99 to 1.62) 0 II　 　Tertiary referralNo Ref.Yes 1.48 (0.78 to 2.83) 1.20 (0.70 to 2.05) 2 II　 　SmokingNever Ref.ever 1.20 (0.94 to 1.52) 1.20 (0.94 to 1.52) 0 I　 　Alcohol use

Never Ref.ever 0.78 (0.59 to 1.01) 0.78 (0.59 to 1.01) 0 II　 　Diabetes mellitusNo Ref.Yes 1.21 (0.63 to 2.32) 1.21 (0.63 to 2.32) 0 II　 　ObesityNoYes 1.24 (0.95 to 1.63) 1.13 (0.84 to 1.53) 1 II　 　Preoperative seton drainage

No Ref.Yes 1.05 (0.51 to 2.16) 1.05 (0.51 to 2.16) 0 II

Non-significant fistula/surgery-related factorsSurgical procedureFistulotomy Ref.Fistulectomy 1.41 (0.73 to 2.73) 1.41 (0.73 to 2.73) 0 IAdvancement flap 1.39 (0.22 to 8.71) 1.39 (0.22 to 8.71) 0 III

Height of internal opening

Low Ref.High 2.75 (0.81 to 9.38) 2.75 (0.81 to 9.38) 0 II

Location of internal opening

Posterior Ref.

Lateral 0.98 (0.57 to 1.69) 0.98 (0.57 to 1.69) 0 IIAnterior 0.76 (0.41 to 1.39) 0.76 (0.41 to 1.39) 0 II

Supralevator extensionsNo Ref.Yes 1.79 (0.62 to 5.21) 1.79 (0.62 to 5.21) 0 III

Postoperative drainageNo Ref.Yes 1.02 (0.78 to 1.32) 1.02 (0.78 to 1.32) 0 II

Type of fistula (fistula classification)Suprasphincteric Ref.Intersphincteric 0.24 (0.02 to 2.40) 0.09 (0.01 to 0.92) 1 IIILow transsphincteric 0.16 (0.04 to 0.67) 0.16 (0.04 to 0.67) 0 IIHigh transsphincteric 0.45 (0.19 to 1.07) 0.45 (0.19 to 1.07) 0 IIAbbreviations: NA, not available; Ref., Reference group; RR, relative risk; T-F, trim and filled method.

eTable 10. Risk Factors of Anal Fistula Recurrence in Patients with Mucosal Advancement Flap Surgery.

Mucosal advancement flap

Risk Factor HeterogeneityNo. of Studies No. of Patients Postoperative

Recurrence RR (95% I2, % P Value

CI)Gender 　

Female Ref. Ref.Male 3 577 0.90 (0.64 to 1.27) 0 0.580

Age (years) 　>40 or 45 Ref. Ref.

<40 or 45 3 577 1.37 (1.01 to 1.87) 0 0.671

Smoking useNever Ref. Ref.ever 3 577 1.25 (0.81 to 1.93) 38.7 0.195

Alcohol useNever Ref. Ref.ever 2 313 0.78 (0.59 to 1.01) 0 0.789

Obesity 　No Ref.Yes 4 638 1.24 (0.95 to 1.63) 49.6 0.114

Prior surgery 　No Ref. Ref.Yes 4 638 1.08 (0.65 to 1.77) 42.9 0.154

Seton historyNo Ref. Ref.Yes 3 418 0.73 (0.47 to 1.12) 12.9 0.317

Horseshoe 　

extensionNo Ref. Ref.Yes 2 357 1.68 (1.21 to 2.34) 0 0.579

Location of internal opening 　Posterior Ref. Ref. 　Lateral 1 105 1.04 (0.44 to 2.49) NA NAAnterior 2 357 0.87 (0.61 to 1.24) 0 0.937

Postoperative drainageNo Ref. Ref.Yes 2 357 1.02 (0.78 to 1.32) 0 0.382

Abbreviations: NA, not available; Ref., Reference group; RR, relative risk.

eTable 11. Sensitivity Analysis for Outcomes of Anal Fistula Recurrence Associated with Patient-related Factors, Fistula and Surgery-related factors.

Study omitted Pooled relative risk (95% CI)Heterogeneity estimate,I2, % (P value)

Patient-related FactorsAgeYano et al 1.42(1.12,1.79) 0%,0.557van Koperen et al (FG) 1.29(0.97,1.73) 59.6%,0.03van Koperen et al (RAFG) 1.25(0.95,1.63) 55.80%,0.046Hamadani et al 1.05(0.93,1.92) 10.40%,0.349Abbas et al 1.30(0.99,1.70) 60.90%,0.025van Onkelen et al 1.30(0.96,1.77) 58.70%,0.033

Zimmerman et al 1.21(0.94,1.57) 50.30%,0.074overall 1.27(0.99,1.62) 53.20%,0.046

SmokingLi et al 1.20(0.91,1.57) 20.80%,0271Devaraj et al 1.09(0.86,1.37) 0%,0.529van Koperen et al (FG) 1.20(0.92,1.57) 20.20%,0.275van Koperen et al (RAFG) 1.22(0.92,1.61) 21.90%,0.262Placer Galán et al 1.23(0.95,1.59) 16.10%,0.307Hamadani et al 1.25(0.99,1.57) 0%,0.427van Onkelen 1.34(1.00,1.80) 0.20%,0.422Zimmerman et al 1.10(0.88,1.39) 0%,0.545overall 1.20(0.94,1.52) 8.9%,0.361

ObesityBoenicke et al 1.29(0.79,2.12) 66.30%,0.052van Onkelen et al 1.36(1.05,1.77) 29.90%,0.24Zimmerman et al 1.24(0.87,1.77) 65.90%,0.053Schwandner et al 1.15(0.94,1.41) 19.50%,0.289overall 1.24(0.95,1.63) 49.6%,0.144

Prior anal surgeryVisscher et al 1.45(0.96,2.19) 54.10%,0.02Li et al 1.39(0.94,2.06) 48.10%,0.044van Koperen et al (FG) 1.53(1.02,2.30) 58.50%,0.01van Koperen et al (RAFG) 1.54(1.02,2.32) 58.50%,0.01Garcia-Aguilar et al 1.48(0.98,2.23) 56.20%,0.015Poon et al 1.42(0.96,2.10) 51.90%,0.028Placer Galán et al 1.50(1.01,2.22) 57.50%,0.012Abbas et al 1.66(1.13,2.43) 50.90%,0.032Boenicke et al 1.44(0.99,2.09) 52.00%,0.027

Zimmerman et al 1.70(1.13,2.54) 49.00%,0.039van Onkelen 1.66(1.05,2.61) 51.10%,0.031overall 1.52(1.04,2.23) 53.9%,0.017

Study omitted Pooled relative risk(95% CI)Heterogeneity estimate,I2, % (P value)

Fistula and Surgery-related factorsLi et al 2.75(0.60,12.67) 62.20%,0.072Jordán et al 1.95(0.95,4.02) 6.60%,0.343Poon et al 3.36(1.06,10.69) 65.40%,0.055Garcia-Aguilar 4.89(1.99,12.01) 0%,0.371overall 2.98(1.10,8.06) 50.30%,0.11

Fistula classificationPlacer Galán et al 4.98(3.84,6.44) 5.70%,0.374Li et al 4.39(3.46,5.57) 0%,0.842Jordán et al 4.98(3.74,6.62) 11.80%,0.338Poon et al 5.05(3.76,6.78) 13.40%,0.328Abbas et al 6.60(4.29,10.16) 0%,0.798Boenicke et al 4.66(3.71,5.85) 0%,0.413overall 4.77(3.87,5.95) 0%,0.463

Internal opening identificationGarcia-Aguilar et al 8.83(3.89,20.04) 0%,0.49Poon et al 7.97(4.56,13.92) 0%,0.61Jordán et al 8.92(5.40,14.71) 0%,0.705overall 8.54(5.29,13.80) 0%,0.784

Height of internal openingMijnsbrugge et al 5.24(2.37,11.59) 0%,0.848Li et al 1.77(0.54,5.66) 56.90%,0.128

Poon et al 2.39(0.54,10.59) 90.70%,0.001overall 2.75(0.81,9.38) 84.50%,0.002

Horseshoe extensionsGarcia-Aguilar et al 1.87(1.29,2.73) 16.10%,0.311Emile et al 1.81(1.34,2.44) 0%,0.719Abbas et al 1.96(1.36,2.84) 25.30%,0.26van Onkelen et al 2.20(1.31,3.68) 24.00%,0.267Zimmerman et al 2.11(1.50,2.97) 0.60%,0.389overall 1.92(1.43,2.59) 3.10%,0.389

No.of fistula tractsAbbas et al 5.79(1.25,26.75) 50.50%,0.155Schulze et al 2.03(0.48,8.60) 0%,0.901Parthasarathi et al 6.08(1.55,23.92) 48.50%,0.164overall 4.77(1.46,15.51) 42.50%,0.176

eTable 12. Sensitivity Analysis for Risk Factors of Anal Fistula Recurrence in Patients with Mucosal Advancement Flap Surgery and Class of Evidence.

Significant factors Postoperative Recurrence RR (95% CI)

T-F adjusted RR (95% CI)

Filled studies Class of Evidence

GenderFemale Ref.Male 0.90 (0.64 to 1.27) 0.77 (0.58 to 1.02) 2 II

Age (years)>40 or 45 Ref.<40 or 45 1.37 (1.01 to 1.87) 1.20 (0.93 to 1.54) 2 II

Smoking useNever Ref.ever 1.25 (0.81 to 1.93) 1.25 (0.81 to 1.93) 0 II

Alcohol useNever Ref.ever 0.78 (0.59 to 1.01) 0.78 (0.59 to 1.01) 0 NA

ObesityNoYes 1.24 (0.95 to 1.63) 1.13 (0.84 to 1.53) 1 II

Prior anal surgeryNo Ref.Yes 1.08 (0.65 to 1.77) 0.92 (0.54 to 1.58) 2 II

Seton historyNo Ref.Yes 0.73 (0.47 to 1.12) 0.73 (0.47 to 1.12) 0 II

Horseshoe extensionNo Ref.Yes 1.68 (1.21 to 2.34) 1.68 (1.21 to 2.34) 0 NA

Location of internal openingPosterior Ref.Lateral 1.04 (0.44 to 2.49) NA NAAnterior 0.87 (0.61 to 1.24) 0.86 (0.62 to 1.19) 1 NA

Postoperative drainageNo Ref.

Yes 1.02 (0.78 to 1.32) 1.02 (0.78 to 1.32) 0 NAAbbreviations: NA, not available; Ref., Reference group; RR, relative risk; T-F, trim and filled method.

eFigure 1. Forest Plot for Pooled Rate of Anal Fistula Recurrence.