| Methods |
Randomized controlled trial |
| Participants |
Patients with chronic anal fissure |
| Interventions |
Anal stretch (78) versus lateral subcutaneous sphincterotomy |
| Outcomes |
Persistence of the fissure |
| Notes |
|
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Random sequence generation (selection bias) |
Low risk |
pulling classification cards once surgery had been determined to be necessary |
| Other bias |
High risk |
‐ insufficient length of follow‐up: only 31 of 156 patients evaluated at 12 months
‐ confounding variable: different anaesthetic for different surgical groups
‐ selection criteria: 27% of patients did not present with anal pain on defecation
‐ also, cases in which healing of the fissure did not occur had significantly elevated sphincter pressure in 14 of 17 cases, denoting incomplete sphincterotomy |
| Blinding of outcome assessment (detection bias)
All outcomes |
Low risk |
physicians examining for recurrence / complications were blinded to patient operative classification |
| Drop‐outs
All outcomes |
High risk |
drop out rate 24% |
