Peck 1982.

Methods	Parallel design, randomised controlled trial
Participants	33 participants from a single centre, the Clinical Center of the National Institutes of Health, at Bethesda, United States Inclusion criteria: having at least ten inflamed deep derma or subcutaneous acne cysts or nodules of at least 4 mm diameter; history of minimal response to treatment with oral and topical antibiotics, oral vitamin A, topical vitamin A acid, topical benzoyl peroxide, x‐irradiation, oral contraceptives, oral dapsone, intralesional injections of corticosteroids, oral prednisone, surgical drainage, applications of liquid nitrogen, photochemotherapy with psoralen and long wave ultraviolet light, and other acne therapies Exclusion criteria: pregnant women and fertile women who refused to use birth control measures Age, gender and duration of acne: not provided Acne severity: not stated if moderate or severe, only reported as cystic acne
Interventions	Group 1 (n = 16): oral isotretinoin at a minimum of 0.5 mg/kg/day for a minimum period of 4 months in the first double‐blind phase of the study. Additional courses of isotretinoin could be given after an 8‐week rest period between courses if it was thought that the participant could benefit from further treatment Group 2 (n = 17): placebo for a minimum period of 1 month At the monthly visit, "if there had been no change or only a slight worsening of the acne, then the dose of isotretinoin or placebo was increased by 0.5 mg/kg/day. If there had been a marked worsening, the protocol permitted a cross‐over to active drug if the patient had been taking placebo."
Outcomes	Primary efficacy outcome: Improvement in the number of cysts assessed by lesion counts Secondary outcomes: Improvement in the number of papules and pustules assessed by lesion count (lesions were counted at baseline, monthly until 15 months, and at the follow‐up evaluation, after 40 months from the start of the therapy) Reduction in sebum production assessed by quantitative analyses of samples obtained according to the method described previously by Strauss and Pochi (Strauss 1961) Clinical and laboratory side effects evaluated at baseline and monthly intervals during and after treatment* * Indicates outcomes which matched those prespecified for this review
Funding body	Hoffman‐La Roche Inc.
Notes	We considered for our review only the first four weeks of the study: after this time point, switching of participants from placebo group to isotretinoin group had started.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The patients were randomly assigned... according to a computer‐generated code." Comment: The trial described an adequate method for the random sequence generation
Allocation concealment (selection bias)	Unclear risk	Comment: There was no statement regarding methods of allocation concealment in the study
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: “If there had been a marked worsening, according to predetermined criteria, then the double‐blind code was broken.” "...or an equivalent number of placebo capsules identical in appearance. The capsules were dispensed to the patients by a third party ..." Comment: Despite having described efforts to keep blinding of participants and personnel during the first study phase ("double‐blind"), there was no explicit reporting of who was and was not blinded across participants, personnel, and outcome assessors
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: It was not clear in the report whether the person who had done the outcome assessment was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: An intention‐to‐treat analysis was probably done, as there was no dropout for the first four weeks of the study which were considered for our analysis (after this time point, switching of participants from placebo group to isotretinoin group had started)
Selective reporting (reporting bias)	High risk	Quote: “In four matched pairs, the average sebum production was 0.25 mg lipids/10 cm2/3 hr, with a range of 0.19 to 0.35, taken from 12 to 16 weeks on therapy”. Comment: No protocol available and the report in the results section suggested that not all participants from each group had been assessed for one outcome (reduction in sebum production); selective choice of subsets of data for this outcome, which was listed in the methods section, might have occurred
Other bias	Unclear risk	Comment: The study might be at risk of inappropriate influence of funders, as a pharmaceutical company promoted it