Keipert 1976.

Methods	Double‐blind, cross‐over randomised controlled trial, 3 weeks then cross‐over to other intervention. Intention‐to‐treat analysis not reported
Participants	19 participants completed. Inclusion criteria: infants with typical clinical changes of seborrhoeic dermatitis of infancy, rash in nappy area, and other typical sites Exclusion criteria: not stated Number of participants randomised: not stated; 19 completed trial; 7 completed 1 extra 3‐week course Number of dropouts: not stated Sex: not stated Mean age: not stated; range: 6 weeks to 6 months Location: Australia, probably secondary care
Interventions	Treatment (n not reported, possibly 10) Oral biotin powder 2mg twice daily for 3 weeks, and topical therapy of betamethasone 12‐valerate cream 0.02% Comparator (n not reported, possibly 10) Placebo powder twice daily for 3 weeks, and topical therapy of betamethasone 12‐valerate cream 0.02%
Outcomes	Primary outcome measures Quantitative change in area of skin involvement measured according to % increase or decrease. Qualitative changes scored using 5 grades of 0–10 points, giving a score out of 50. Not reported what the qualitative criteria measured were. Actual scores and percentages not reported Assessment performed at end of each 3‐week period, with no reported washout time. No reporting of adverse effects. No reporting of quality of life outcomes.
Notes	Note: 7 participants reported to have completed 3 courses of treatment, creating a total of 45 courses of treatment. Study design flawed given that the condition was likely to improve rapidly over time especially given that both groups received a topical steroid that was likely to improve the rash.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "independently allotted at random". Comment: did not specify method of random sequence generation.
Allocation concealment (selection bias)	Unclear risk	Quote: "independently allotted at random". Comment: no further details given.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	No description of presentation of the 2 interventions.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "Changes were blindly assessed at the end of each three‐week period". Comment: no further details given.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Not reported how many participants were recruited.
Selective reporting (reporting bias)	Unclear risk	No study protocol.
Other bias	High risk	No statistical methods reported. Pharmaceutical company provided the products and provided help with literature and statistical analysis.