Reconsidering racial/ethnic differences in sterilization in the United States.

OBJECTIVE Cross-sectional studies have found that low-income and racial/ethnic minority women are more likely to use female sterilization and less likely to rely on a partner's vasectomy than women with higher incomes and whites. However, studies of pregnant and postpartum women report that racial/ethnic minorities, particularly low-income minority women, face greater barriers in obtaining a sterilization than do whites and those with higher incomes. In this paper, we address this apparent contradiction by examining the likelihood a woman gets a sterilization following each delivery, which removes from the comparison any difference in the number of births she has experienced. STUDY DESIGN Using the 2006-2010 National Survey of Family Growth, we fit multivariable-adjusted logistic and Cox regression models to estimate odds ratios and hazard ratios for getting a postpartum or interval sterilization, respectively, according to race/ethnicity and insurance status. RESULTS Women's chances of obtaining a sterilization varied by both race/ethnicity and insurance. Among women with Medicaid, whites were more likely to use female sterilization than African Americans and Latinas. Privately insured whites were more likely to rely on vasectomy than African Americans and Latinas, but among women with Medicaid-paid deliveries reliance on vasectomy was low for all racial/ethnic groups. CONCLUSIONS Low-income racial/ethnic minority women are less likely to undergo sterilization following delivery compared to low-income whites and privately insured women of similar parities. This could result from unique barriers to obtaining permanent contraception and could expose women to the risk of future unintended pregnancies. IMPLICATIONS Low-income minorities are less likely to undergo sterilization than low-income whites and privately insured minorities, which may result from barriers to obtaining permanent contraception, and exposes women to unintended pregnancies.