Predictors of contraceptive discontinuation in a sexually transmitted disease clinic population.

CONTEXT Women who attend sexually transmitted disease (STD) clinics are at high risk for unintended pregnancy. Little information is available, however, on the rates of discontinuation of effective contraceptive method use among this population. METHODS As part of a study on contraceptive services offered by an STD clinic in Denver, 406 clients who accepted these services in 1996-1999 were interviewed about their contraceptive practice, experience of side effects and method-use problems at baseline and at four, eight and 12 months of follow-up. Multivariate survival analysis was used to assess predictors of discontinuation of effective contraceptive use. RESULTS Twenty-nine percent of women discontinued use by the end of one year. Coxproportional hazards models show that compared with women who reported no method-use problems, those who experienced one problem were three times as likely (hazard ratio, 3.0) to discontinue effective use, and women who had at least two problems were five times as likely (5.0) to discontinue use. The experience of side effects with either a past or a current method, however, was not associated with the risk of discontinuation. Furthermore, women who reported risky sexual behavior in the year before enrollment were significantly less likely to discontinue effective method use (hazard ratio, 0.4), as were women who were covered by medical insurance or who gained such coverage during a follow-up interval (hazard ratio, 0.5 for each). CONCLUSIONS In this study population of STD clinic users, method-use problems appear to be a more fundamental issue for contraceptive compliance than the pastor current experience of side effects. The unexpected association between method-use problems and the risk of discontinuation needs to be further delineated so that effective interventions addressing these problems can be developed and implemented.