Background: The progestogen component of combined oral contraceptives (COC) has undergone changes since it was first recognised that it's chemical structure could influence the spectrum of minor adverse and beneficial effects. The major determinants of effectiveness are compliance and continuation which may be influenced by cycle control and common side effects. The rationale of this review is to provide a systematic comparison of COCs containing the progestogens currently in use worldwide.

Objectives: To compare currently available low-dose COCs containing ethinyl estradiol and different progestogens in terms of contraceptive effectiveness, cycle control, side effects and continuation rates.

Search strategy: A search of PubMed, LILACS, EMBASE, Popline, CINAHL and the Cochrane Central Register of Controlled Trials databases was conducted in September 2010 to update the 2004 review.

Selection criteria: Randomised trials reporting clinical outcomes were considered for inclusion. We excluded studies comparing monophasic with multiphasic pills, crossover trials, trials in which the difference in total content of ethinyl estradiol between preparations exceeded 105 µg per cycle and those comparing continuous dosing regimens.

Data collection and analysis: Two reviewers independently assessed methodological quality, applied inclusion criteria and extracted data.

Main results: Thirty trials with a total of 13,923 participants were included, generating 16 comparisons. Overall the quality of trials was low. Only four trials were double-blind. At least twenty-three trials were sponsored by pharmaceutical companies. There was less discontinuation with second-generation compared with first-generation monophasic progestogens (3 trials, 2,709 women, Relative Risk (RR) 0.76, 95% Confidence Interval (CI) 0.67-0.86); this remained significant when only double-blind trials were considered (812 women, RR 0.79, 95% CI 0.66-0.94).Women using monophasic COC's containing third-generation progestogens were less likely to discontinue than the second-generation group (3 trials, 1,815 women, RR 0.77, 95% CI 0.60-0.98) but this was not significant when only double-blind trials were considered (RR 0.79, 95% CI 0.50-1.26]. Women in the third-generation group experienced less intermenstrual bleeding than the second-generation group (one double-blind trial, 456 women, RR 0.71, 95% CI 0.55-0.91).Compared to desogestrel (DSG), women in the drospirenone (DRSP) group were more likely to complain of breast tenderness (5 trials, 4,258 women, RR 1.39, 95% CI 1.04-1.86) and nausea (6 trials, 4,701 women, RR 1.46, 95% CI 0.96-2.21].Pregnancy rates overall were comparable but the trials had insufficient power to find potentially important differences.

Authors' conclusions: Women using COCs containing second-generation progestogens may be less likely to discontinue than those using COCs containing first-generation progestogens. Based on one small double-blind trial, third-generation progestogens may be preferable to second-generation preparations with regard to bleeding patterns but further evidence is needed. Without blinding as to treatment group, comparisons between the various "generations" of progestogens used in COCs cannot be made. Until this widespread methodological flaw is overcome in better trials conducted according to CONSORT guidelines and internationally accepted definitions, no further conclusions can be drawn.