Abstract
OBJECTIVE: This paper seeks to determine factors associated with nonuse of contraception by women at risk of unintended pregnancy in the United States. This nonuse may be associated with about 900,000 unintended births in the US each year. STUDY DESIGN: The 2002 and 2006-2010 National Surveys of Family Growth were combined to yield a nationally representative sample of 9,445 women at risk of unintended pregnancy. Logistic regression analyses identified factors associated with nonuse of contraception. RESULTS: This analysis reveals previously undocumented patterns of nonuse: controlling for confounding variables, cohabiting women [adjusted odds ratio (AOR)=2.3, 95% confidence interval (CI)=1.45-3.52] had higher odds of nonuse than married women; women who reported a difficulty getting pregnant (AOR=2.5, 95% CI=2.01-3.01) had higher odds of nonuse than those who did not. Nonuse was also more common among women with a master's degree or more (AOR=1.5, 95% CI=1.11-2.08) compared with those with some college or bachelor's degree, and it was more common among women in their first year after first intercourse than after the first year (AOR 1.6, 95% CI=1.12-2.22). Among women who had a recent unintended birth, the most common reason for not using contraception prior to conception was that she did not think she could get pregnant. CONCLUSIONS: This study establishes national estimates of reasons for nonuse of contraception and identifies some new subgroups at risk of nonuse. IMPLICATIONS: These results may help better understand factors affecting nonuse of contraception and develop strategies for preventing unintended pregnancy in the United States.