Study	Reason for exclusion
Barranco 1974	Not a randomized controlled trial.
Carmina 2002	Used subset of women from two trials: Lucky 1997 and Redmond 1997.
Chapdelaine 1989	Insufficient acne data reported.
Coney 2001	Report focused on weight as outcome, although data were from a trial that looked at acne originally.
Cremoncini 1976	Not a randomized controlled trial.
Erdmann 1994	Not a randomized controlled trial.
Erkkola 1990	Insufficient acne data reported.
Greenwood 1985	Insufficient acne data reported.
Gruber 1998	Insufficient acne data reported.
Grund 1975	Not a randomized controlled trial.
Huber 2000	Insufficient acne data reported.
Katz 2000	Insufficient acne data reported.
Lachnit‐Fixson 1979	Insufficient acne data reported.
Mehrl 1988	News article reporting on two trials. Insufficient acne data reported.
Miller 1986	Insufficient acne data reported.
Nilsson 1967	Insufficient acne data reported.
Perrone 1987	Not a randomized controlled trial.
Sanam 2011	Sample sizes for analysis not provided for acne data. Unable to obtain further information from researcher.
Sanhueza 1979	Insufficient acne data reported.
Spona 1996	Insufficient acne data reported.
Vegetti 1996	Insufficient acne data reported.
Vermeulen 1988	Acne was not outcome.
Vexiau 2002	Insufficient acne data reported (acne was grouped with hirsutism as an assessment of 'hyperandrogenism'). Treatments were a COC plus topical minoxidil versus a COC containing cyproterone acetate/ethinyl estradiol plus cyproterone acetate supplement.
Volpe 1994	Not a randomized controlled trial.
Weber‐Diehl 1993	Insufficient acne data reported.
Wendler 1995	Randomization not implemented; group assignment affected by acne.
Wishart 1991	Not a randomized controlled trial.