Iselin 1990.

Methods	2‐arm quasi‐randomised controlled trial. Contaminated (class 3) surgery
Participants	268 consecutive patients presenting for urgent hand surgery were enrolled. 45 participants were excluded after randomisation; 5 were included by mistake, 33 were lost to follow‐up, 5 were excluded for protocol violation, 2 for unknown cause. Exclusion criteria: patients currently receiving immunosuppressants including steroids, renal or hepatic disease, allergy to iodine or rifampicin, burns, cutaneous laceration only, iodising radiation, current malnutrition Setting: hospital inpatients and outpatients
Interventions	Intervention: before closure wound was rinsed with rifamycin solution and after closure wound was covered with soaked pad of rifamycin. Control: before closure wound was rinsed with povidone iodine dermal solution and after closure wound was covered with soaked pad of povidone iodine dermal solution.
Outcomes	SSI (dichotomous); adverse effect (dichotomous); wound healing (dichotomous)
Notes	45 participants were excluded after randomisation; 5 were included by mistake, 33 were lost to follow‐up, 5 were excluded for protocol violation, 2 for unknown cause. Definition of SSI: presence of clinical criteria for wound infection Definition of adverse effects: cutaneous intolerance, need for further surgery, unspecified general complications Definition of wound healing ‐ proportion of wounds healed: clinical criteria for healing at 1 week and 2 weeks Concurrent Illness: none reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Comment: this study was quasi randomized, alternate patients were randomized into 2 parallel groups.
Allocation concealment (selection bias)	High risk	Comment: allocation concealment was not described, however alternate allocation would allow prediction of allocation sequence.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: participants and personnel were unable to be blinded, as the 2 interventions were different in appearance.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: blinding of outcome assessor was not reported.
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: 45 participants were excluded after randomisation, this accounts for 17% of enrolled participants. 20 missing participants were from the intervention group and 25 were from the control group. 5 were included by mistake, 33 were lost to follow‐up, 5 were excluded for protocol violation, 2 for unknown cause.The analysis was made per protocol rather than ITT.
Selective reporting (reporting bias)	Unclear risk	Comment: outcome data were presented for all outcomes listed in methods. Study protocol was not available to identify any unreported outcomes. Patients excluded from the trial were well recorded, as were patients who dropped out of the trial.
Other bias	Low risk	Comment: none were identified