Tang 2015.

Methods	Country where data collected: China Parallel‐group RCT Unit of randomisation: participant Unit of analysis: unclear (burn?) Duration: 4 weeks
Participants	Inclusion criteria: deep partial‐thickness thermal burn injury covering 2.5%‐25% TBSA (third‐degree areas were not to exceed 10% TBSA). aged 5‐65 years; at least one isolated burn area not on head or face with deep partial 2nd‐degree burn from 1%‐10% TBSA Exclusion criteria: burns older than 36 h, clinically infected; treated with active agent before study entry (SSD allowed up to 24 h before randomisation); dermatologic disorders or necrotising processes; underlying diseases such as HIV/AIDS, cancer, severe anaemia, insulin‐dependent diabetes, systemic glucocorticoid use except occasional prednisolone < 10 mg/d; immunosuppressive agents, radiation or chemotherapy in previous 30 days; known allergy/sensitivity to the products; pregnancy; previous participation in this (or other study within 1 month) Participants: 158 randomised participants (total number of burns > 200) Mean age (years): 36.2 (range 5.2‐65.5; only 5 < 12 years). No difference between groups Male participants: 55 vs 57 Burn type: scald 30 vs 41; flash 8 vs 7; flame 32 vs 31; contact 1 vs 3 Burn degree & TBSA: 2nd‐degree superficial partial 4.48% vs 4.29%; deep partial‐thickness 6.28% vs 5.18%; third‐degree 0.345% vs 0.317%. enrolled study site: 2.72 vs 2.64 Burn location: arm 52 vs 53, buttock 6 vs 7, hand 41 vs 42, leg 29 vs 35, thigh 24 vs 31, trunk 26 vs 27, other 43 vs 46
Interventions	Intervention arm 1: absorbent foam silver dressing (Mepilex Ag) changed every 5‐7 days; gauze wrap as secondary dressing. N = 73 Intervention arm 2: SSD 1% cream; gauze pad and wrap as secondary dressing. N = 85 Cointerventions: debrided and/or cleansed according to standard practice
Outcomes	Primary outcome: wound healing primary outcome: infection Secondary outcome: adverse event Secondary outcome: pain
Notes	Funding NR
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Enrolled subjects were assigned randomly using a block design, with block sizes varying between 2.4 and 6 (in Viedoc, Pharma Consulting Group, Uppsala, Sweden) to either SSD or Mepilex Ag. Subjects were consecutively allocated to the treatment at each center and given a subject code, depending on which strata they belonged to." Comment: randomisation sequence computer‐generated using blocking design
Allocation concealment (selection bias)	Low risk	Quote: "Enrolled subjects were assigned randomly using a block design, with block sizes varying between 2.4 and 6 (in Viedoc, Pharma Consulting Group, Uppsala, Sweden) to either SSD or Mepilex Ag. Subjects were consecutively allocated to the treatment at each center and given a subject code, depending on which strata they belonged to." Comment: allocation conducted remotely by consecutive allocation of codes within stratified design
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "In addition, the investigator was required to make a subjective assessment of healing at each weekly assessment before cleansing and/or debridement. Percentage of the burn healed since baseline was to be performed by a blinded observer." Comment: assessment of healing was conducted by an assessor blinded to the treatment allocation; it's not clear whether assessment of other outcomes was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: ".....158 patients were randomized, and 153 patients were subjected to at least one treatment and were included in the ITT population, 71 (46%) of them randomized to Mepilex Ag and 82 (54%) randomized to SSD. Thirteen patients (8%) discontinued before the study ended, 5 (7%) of them from the Mepilex Ag group and 8 (10%) from the SSD group. One patient withdrew consent from the SSD group, while the other 12 discontinued because of other reasons (Fig. 1)." Comment: all participants were accounted for and the proportion who discontinued was low and low relative to the event rate for healing
Selective reporting (reporting bias)	Low risk	Quote: "The primary end point was time to healing (defined as 95% epithelialisation by visual inspection). The secondary end points were percentage of burns epithelialised/healed, numbers of burns healed or not at each visit (not at baseline), number of study burns requiring a skin graft, and number of dressing changes. Additional outcomes were measured assessing the tolerability and performance of the dressings on wound and periwound status, including pain and experience of use of the dressings." Comment: the defined outcomes were all fully reported
Other bias	Unclear risk	Comment: it was unclear how the designated burn was chosen in participants with multiple burns. However it was clear that there were no unit of analysis issues.