Refining assessment of contraceptive use in the past year in relation to risk of unintended pregnancy.

<bold>Objectives: </bold>Healthy People 2020 establishes objectives for reducing the proportion of pregnancies in the United States that are unintended and for improving contraceptive use. This analysis describes ways to more closely align measurement of contraceptive use with periods of risk for unintended pregnancy using the National Survey of Family Growth (NSFG).<bold>Methods: </bold>Using the 2011-2015 NSFG we constructed two measures of contraceptive use for women we defined as at risk of an unintended pregnancy: (1) we augmented a measure of recent contraceptive use by recoding non-users according to their method use during their last month of sex in the past 12 months; (2) we augmented use at last sex in the past 12 months by excluding women who were pregnant at last sex. Estimates were compared overall and within 5-year age groups.<bold>Results: </bold>The augmented measure of recent contraceptive use found fewer women to be using no contraception than the standard measure (7.3% vs 15.4%; p < .001); greater differences were found between the two measures for younger women. When considering contraceptive use at last sex, the augmented measure identified fewer women as using no contraception (15.8% vs 21.0%; p < .001) than the standard measure and more women to be using a most effective method (33.3% vs 31.1%; p = .04) than the standard measure.<bold>Conclusions: </bold>Aligning periods of unintended pregnancy risk with contraceptive use assessment reduced estimates of no contraceptive use; changes in estimates by method type varied by age.<bold>Implications: </bold>When assessing contraceptive use for the purpose of unintended pregnancy prevention, researchers may consider the methods described here to further align contraceptive use measurement with periods of unintended pregnancy risk. [ABSTRACT FROM AUTHOR]