Ozaki 2015.

Methods	2‐arm RCT undertaken in the USA (2 medical centres)
Participants	500 adults were randomised. Eligible participants were undergoing an open (incision below the inguinal ligament) non‐emergency surgical procedure for peripheral vascular disease involving arteries or bypass grafts, with the anticipation that all incisions would be closed.
Interventions	Group A (n = 250): standard gauze Group B (n = 250): silver alginate dressings No other dressing details provided. This original operating room dressing remained in situ until gross soiling, clinical need to remove, or postoperative day 3, whichever came first. Subsequent care was at the provider’s discretion. The wound‐closure technique was at the discretion of the surgeon. Cyanoacrylate tissue adhesives were considered as dressings and were not permitted.
Outcomes	Primary review outcome: assessment of wound complication which included SSI at 30 days (defined as no wound complication, superficial SSI or deep SSI ‐ noted that National Surgical Quality Improvement Program definitions were used). Secondary review outcomes: none reported
Notes	Trial outcome data: see Table 12 Noted that most cases were clean surgery (notes 25/500 participants had wounds classified as clean/contaminated) Follow‐up: 30 days We contacted the author to ask about methods
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomised in the operating room by block design after wound closure was completed but before any dressing was applied." Quote from author: "Block 16 randomization per site. Generated via RAND in SAS" Comment: adequate
Allocation concealment (selection bias)	Low risk	Quote from author: "Patients and providers were blinded until time of reveal, which was at the end of the case (patient under anaesthesia) when the dressing is needed." Comment: adequate
Blinding (performance bias and detection bias) All outcomes	High risk	All outcomes Quote from paper: "In addition, although the patients and providers were not formally blinded to the type of original postoperative dressing, the study physicians generally reported an inability to recall which dressing the patient had received at the late follow‐up visits." Quote from author: "No one [sic] who assessed outcome was formally blinded, though in reality the evaluating clinicians noted that they frequently did not recall which early post‐operative dressing the patient had at 2 and 4 weeks." Comment: blinded outcome assessment was not conducted
Incomplete outcome data (attrition bias) All outcomes	Low risk	In total 7/500 participants (3 in the silver group and 4 in the gauze group) were lost to follow up.
Selective reporting (reporting bias)	Unclear risk	Study reported and listed a number of other outcomes that were collected, including length of stay and EQ‐5D, but these data were not reported in the paper.
Other bias	Low risk	None noted