Using Longitudinal Data to Understand Changes In Consistent Contraceptive Use

CONTEXT: Most studies of contraceptive behavior rely on cross-sectional data and are unable to adequately measure fluctuations in contraceptive use or changes in circumstances and attitudes that are likely to be associated with this outcome. METHODS: Between November 2012 and May 2014, four waves of data were gathered from a national sample of 1,842 women aged 18-39 at baseline. Cross-tabulations were used to examine change and stability in time-varying characteristics theorized to be associated with consistent contraceptive use. Random-effects and fixed-effects logistic regression models were used to examine variables associated with consistent contraceptive use. RESULTS: While a majority of women were at risk of unintended pregnancy during each survey period, only 42% were at risk during all four. Random-effects logistic regression analysis revealed that the odds of being a consistent contraceptive user were 10 times as high for a woman who expressed a strong pregnancy avoidance attitude as for a woman who had a weak attitude. This strong association was confirmed in the fixed-effects model. However, having a strong desire to avoid pregnancy was not static; among women at risk of unintended pregnancy during at least one survey period, 53% reported a change in attitude. CONCLUSIONS: These findings build on prior research suggesting that pregnancy avoidance attitudes are an important motivator for contraceptive use. It is critical to recognize that the context in which many women make decisions about pregnancy and contraceptive use changes over relatively short periods of time.