Manizate 2012.

Methods	RCT, parallel; single‐centre trial, legs randomised (within‐participant) Setting: tertiary‐care referral wound practice. Country: USA Duration of intervention: 8 weeks Funding: industry‐funded trial ‐ Medline Industries; Mundelein, Illinois Unit of analysis: ulcer
Participants	10 participants with VLUs (hard‐to‐heal: ulcer size). Number of wounds: 20. 2 per participant randomised to different groups; unclear if ulcers selected. (9/10 participants had VLU; 1 had DFU (apparently all data reported)) Age: not reported. Sex (M/F): not reported. Duration of ulcer: not reported. Ulcer size: Group 1: 14.9 (SD 13.3) cm² versus Group 2: 9.8 (SD 9.7) cm². No infected wounds at baseline (but bacterial loads reported for both groups; no conversion to infection). Bilateral comparable wounds Inclusion criteria: aged > 18 years, full thickness venous stasis or diabetic or neuropathic lower‐extremity wounds, greater or lesser saphenous insufficiency; venous perforator incompetency and deep venous system incompetency or diabetes and HbA1c 6%‐14% and ABI 0.7‐1.2 Exclusion criteria: known history of poor compliance or allergy to products evaluated; NPWT in previous 14 days; skin substitutes or skin grafts in previous 60 days; participants requiring corticosteroids or with immune disorders
Interventions	Group 1: PMM + silver dressing ‐ bovine native collagen plus silver (manufacturer not stated); secondary foam dressing (Optifoam); (n = 10; duration 8 weeks) Group 2: hydrocolloid + silver dressing ‐ carboxymethylcellulose plus silver (manufacturer not stated); secondary foam dressing (Optifoam); (n = 10; duration 8 weeks) Co‐interventions: 4‐layer multilayer wrap for compression (4‐Layer Compression Bandaging System) Dressing procedure: sharp debridement; cleansing with normal saline; secondary foam dressing (Optifoam). Dressings were changed weekly. Prior treatment: not reported
Outcomes	Primary outcomes of the review: proportion completely healed (8 weeks ‐ assumed); adverse events not reported Secondary outcomes: infection, change in ulcer size. Pain (general) measured on a pain scale, but no results given
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "1 limb was randomized to treatment with either CMC or BDC, whereas the contralateral wound was treated with the other dressing"
Allocation concealment (selection bias)	High risk	Quote: "1 limb was randomized to treatment with either CMC or BDC, whereas the contralateral wound was treated with the other dressing". Comment: Baseline differences in wound size ‐ mean (SD): BDC 14.9 (SD13.3) and CMC 9.8 (SD 9.7) cm². Additionally, the absolute rate of wound closure was bigger in BDC, but the percentage (of wound volume) rate of closure was smaller in BDC. The difference was not statistically significant, but this was a small study and the differences in an important prognostic factor for healing, together with the lack of information on allocation concealment in a within‐participant trial, suggests high risk of selection bias
Blinding participants and personnel (performance bias)	High risk	Quote: "This is a prospective, randomized, nonblinded trial."
Blinding outcome assessors (detection bias): healing outcomes	Unclear risk	Quote: "At the weekly study, site dressing changes, subjective assessments of .. any signs of erythema (no reddening, pink, red, beet red), the level of pain (linear analog scale 1 through 10) were recorded. Digital images also were taken and used to assess wound healing over time. Moreover, the total surface area (in centimeters squared) of the participant’s reference ulcers was measured." Comment: unclear who the outcome assessors were for healing
Blinding outcome assessors (detection bias): secondary outcomes	High risk	Quote: "This is a prospective, randomized, nonblinded trial." Comment: outcome assessors were the participants for pain
Incomplete outcome data (attrition bias) healing/secondary	Low risk	Comment: no missing data
Selective reporting (reporting bias)	High risk	Quote from the methods section: "...and the level of pain (linear analog scale 1 through10) were recorded. Digital images also were taken and used to assess wound healing over time." Comment: the study authors do not report pain data, although measured. Additionally, the methods section mentions wound healing over time, but no results are reported.
Other bias	High risk	No account taken of paired data. Healing results calculated from percentages and number randomised