Huang 2007.

Methods	Country where data collected: China Parallel‐group RCT (multicentre) Unit of randomisation: participant Unit of analysis: burn Duration: 20 days
Participants	Inclusion criteria: men and women aged 18‐65 years with burn wounds unhealed 3 weeks after injury (residual burn wounds) Exclusion criteria: serious complications of heart, liver, kidney or blood system (blood production or bleeding issues); serious complications, shock or serious systemic infection; uncontrolled diabetes, pregnancy or breast feeding, allergy to solver ions; other reason unable to complete observation period Participants: 111 participants randomised, 98 analysed with 166 burns Mean age (years): NR Male participants: NR Burn type: NR (residual wound) Burn degree: NR Burn size (%TBSA): NR
Interventions	Intervention arm 1: nanocrystalline silver dressing (Acticoat) changed once daily where redness, swelling and high levels of exudate, otherwise every 3 days. Auxilliary dressing over intervention dressing. (83 burns analysed) Intervention arm 2: SSD (5 g per 80 cm2) changed once daily. (83 burns analysed) Cointerventions: washing/rinsing of wounds with sterile water
Outcomes	Primary outcome: wound healing Primary outcome: change in infection status Secondary outcome: adverse events
Notes	Data extracted from English language publication; 2 additional Chinese language publications Funding NR This was a "split‐body" or "intra‐individual" design where a person with two wounds had one wound randomised to each treatment. It was not clear whether the analysis took account of this.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "A multi‐center, randomized experimental design is adopted, with blinding and positive parallel control. The clinical trial was done in four burn centers throughout the country at the same time with the same experimental design. The observing doctor hands out the dressing to every patient according to the time that they come to the hospital and to a randomized serial number." Comment: no information on how the randomisation sequence was derived
Allocation concealment (selection bias)	Unclear risk	Quote: "A multi‐center, randomized experimental design is adopted, with blinding and positive parallel control. The clinical trial was done in four burn centers throughout the country at the same time with the same experimental design. The observing doctor hands out the dressing to every patient according to the time that they come to the hospital and to a randomized serial number." Comment: insufficient information on whether the allocation sequence was adequately concealed
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "Standards for the healing of wound: the wound healed was determined by inspection by two doctors." Comment: no information on whether the outcome assessors were blinded (although the trial is described as blinded)
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "Altogether 111 patients were enrolled in this group, in the process of the trial, 13 patients were dropped out of the study. Among them two patients were dropped out because of silver allergy. Eight were removed because they left to their local clinic before the wound healed, therefore we do not have their related records. Three patients were dropped because of liver dysfunction. The remaining 98 patients who were included in the statistical analysis had altogether 166 residual wounds" Comment: 13/111 participants were not included in the analysis. The event rate was high so although there is potential for differential missing data the impact on the effect estimate was probably small.
Selective reporting (reporting bias)	Unclear risk	Quote: "This study is to investigate the efficacy and safety of nanocrystalline silver (Acticoat) in the treatment of burn wounds, and to assess the clinical value of this dressing." Comment: no specification of how efficacy and safety was to be assessed so difficult to determine if all planned outcomes were reported. However a statistical analysis for wound healing rate was pre‐specified and presented
Other bias	High risk	Comment: unit of analysis issues arising from randomisation at the participant level and analysis at the level of the burn (multiple burns for some participants)