An association between oral contraceptives and venous thrombosis has long been evident; what isn't as clear is which constituent medications-or which amounts-carry the greatest risk. Authors of a recent meta-analysis examined the risk of venous thrombosis according to the type of progestogen, categorizing them according to date of creation as being first-, second-, or third-generation progestogens.

Data from 23 of the 26 studies analyzed pertained to the progressive generations of progestogen: first-generation progestogens included lynestrenol and norethisterone; second-generation progestogens included norgestrel and levonorgestrel; third-generation progestogens included desogestrel, gestodene, and norgestimate. The risk of venous thrombosis was 3.2-times higher with first-generation progestogens, 2.8-times higher with second-generation progestogens, and 3.8-times higher with third-generation progestogens in users of combined oral contraceptives compared with nonusers.

Because the studies themselves didn't compare the progestogens according to generation, the authors also compared 10 frequently prescribed combinations of ethinyl estradiol and progestogens (14 studies). The dose of ethinyl estradiol appeared to play a role: although all 10 combinations were associated with an increase in venous thrombosis risk compared with no use, the risk was highest with 50 micrograms ethinyl estradiol plus levonorgestrel. Risk was also higher with 30 micrograms ethinyl estradiol plus desogestrel, 35 micrograms ethinyl estradiol plus cyproterone acetate (not categorized by generation), and 30 micrograms ethinyl estradiol plus drospirenone (not categorized). Risk was lowest with 30 or 20 micrograms ethinyl estradiol plus levonorgestrel and 20 micrograms ethinyl estradiol plus gestodene. Fifteen of the studies included a nonuser group, and results showed that women who took combined oral contraceptives had an overall risk of venous thrombosis nearly four times that of nonusers.

The authors point out that the levonorgestrel intrauterine device is not associated with increased thrombosis risk, but for women who prefer oral contraceptives, clinicians should prescribe those with the lowest risk and greatest associated adherence.

Stegeman BH, et al. BMJ 2013; 347:f5298.