Sisto 2007.

Methods	Parallel‐group randomised controlled trial, 1 eye per person
Participants	Country: Italy Number of participants (eyes): 40 (40) % women: 35% Average age: 61 years (range 36 to 75) Risk of trabeculectomy failure: high Inclusion criteria: neovascular glaucoma IOP < 21 mmHg and resistant to medical therapy no previous ocular surgery best corrected visual acuity > ‐1.5 logMAR units Exclusion criteria: not reported
Interventions	MMC (22 eyes) Application: intraoperative sponge application Dose: 0.2 mg/ml for 2 minutes Location: between the sclera and the Tenon's capsule 5‐FU (18 eyes) Application: postoperative injections, commencing on day 7, 2 injections per week for 2 weeks and then 1 injection per week for 6 weeks Dose: 0.1 ml of 50 mg/ml Location: subconjunctival injections near the bleb Fornix‐based conjunctival flaps with single surgeon. No releasable sutures or suture lysis employed
Outcomes	Success defined as IOP < 21 mmHg at final postoperative visit. Qualified success defined if IOP < 21 mmHg with addition of topical treatment. Failure is uncontrolled IOP equal or above 21 mmHg or vision dropped to no perception of light. Follow‐up: every 3 months in the first year, every 6 months thereafter to maximum 60 months (5 years)
Notes	Date study conducted: January 1993 to November 2000 Conflict of interest: Not reported Funding source: Not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	40 consecutive people with neovascular glaucoma selected. "All eyes had been assigned with a computer generated randomization code."
Allocation concealment (selection bias)	Unclear risk	No statement made
Blinding (performance bias and detection bias) All outcomes	Low risk	Surgeon not masked due to technique, but the follow‐up staff were masked on collecting postoperative data
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participant data included in analysis. However, length of follow‐up was variable and no statement was made regarding the participants lost to follow‐up.
Selective reporting (reporting bias)	Low risk	Success criteria defined in the methods
Other bias	Unclear risk	No attempted power calculations