Bakri 2003.

Methods	Study design: randomized controlled trial, parallel group Unit of analysis: eyes Number randomized: 201 total eyes, 103 in the fluorouracil group, 98 in the endonasal laser dacryocystorhinostomy (ELDCR) with isotonic saline group Number of arms: 2 Enrollment start year: not reported Length of follow‐up: 12 months or later Sample size calculations: not reported Losses to follow‐up: not reported
Participants	Country: England Age (mean (SD)): 66 (NR) total Females (n (%)): 50 (63) in the ELDCR group, 51 (67) in the ELDCR with isotonic saline group Inclusion criteria: evidence of primary acquired nasolacrimal duct obstruction, symptoms severe enough to require surgery Exclusion criteria: not reported Study group differences: not reported
Interventions	Intervention: ELDCR surgery with application of 0.5 mg/mL FU Comparison intervention: ELDCR surgery with application of isotonic sodium chloride solution
Outcomes	Measured outcomes: functional success, defined as the relief of epiphora anatomic success, defined as patency to lacrimal irrigation ostium size on nasal endoscopy postoperatively Adverse events: none
Identification	Author name: Karim Bakri Institution: University Hospital Email: nick.jones@nottingham.ac.uk
Notes	Funding source: not reported Declarations of interest: not reported Trial registration number: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Authors do not specify how participants were randomized, only that "randomization was performed in the pharmacy department".
Allocation concealment (selection bias)	Low risk	"randomization was performed in the pharmacy department." "Surgeons and patients remained masked to the choice of treatment until the study and follow‐up had been completed", hence we determined that treatment allocation was concealed de facto.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Surgeons and patients remained masked to the choice of treatment until the study and follow‐up had been completed"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Masking of outcome assessors was not reported.
Incomplete outcome data (attrition bias) All outcomes	High risk	Article provides insufficient information regarding time points. Follow‐up time was 12 months or longer, but that could have been anywhere between 12 and 60 months (for the control) or 12 and 48 months (for the intervention). As‐treated analysis was conducted: "This figure may partly reflect a subgroup of patients who did not attend because their symptoms had been relieved. However... these patients were excluded from the statistical analysis".
Selective reporting (reporting bias)	Unclear risk	Study mentions a protocol, but none accessible for comparison with published study. The conclusions state that: "The study sought to determine whether the application of topical fluorouracil reduced scar formation and improved patency rates"; only patency rates (via epiphora) and postoperative levels of fluorouracil were reported in the results.
Other bias	Low risk	Study appears to be free of other sources of bias.