State-level Factors that Affect Post-partum Contraceptive Use Among Low Income Women

Background: Use of contraception postpartum increases the likelihood that any subsequent pregnancy is planned and appropriately spaced. All postpartum women should have easy, affordable, and timely access to contraceptive information and methods. Access for low-income women may vary by state due to differences in publicly-funded family planning (FP) services, the health care infrastructure or the reproductive health (RH) “climate.” Identifying state-level factors that affect postpartum contraceptive use (PP use) is important. Methods: To estimate PP use prevalence and identify maternal characteristics associated with contraceptive use in multivariable modeling, we used data from the 2006-2010 National Survey of Family Growth (NSFG) and 2005-2007 Pregnancy Risk Assessment Monitoring System (PRAMS) surveys in twelve states. Analyses were conducted using two measures of PP use: Effective Method PP use (method failure < 10%) and Any Method PP use (all methods). To identify state-level factors that affect PP use, we conducted a multilevel analysis using the individual-level PRAMS data and various state-level data. While controlling for the maternal characteristics associated with PP use in the multivariable analysis, we assessed the relationship between diverse aspects of the publicly-funded FP system, the health care infrastructure, and the RH climate and PP use. Results: Among low-income, adult women we found a lower prevalence of both Any and Effective Method PP use in the NSFG (79% and 55%) than the twelve-state PRAMS dataset (91% and 72%) or any state separately. Maternal characteristics associated with PP use varied by survey and method effectiveness. Three state-level factors were associated with Any Method (all RH climate-related) and two with Effective Method PP use. Both a smaller proportion of delivery hospitals that are Catholic (OR: 1.36; 95% CI: 1.20-1.54) and more publicly-funded FP clinics per women in need of subsidized services (OR: 1.28; 95% CI: 1.12-1.46) increased the odds of Effective Method PP use. Conclusions: Some of the difference in NSFG and PRAMS prevalence may be an artifact of survey design differences. Some of the variation in PP contraceptive use across states may be due to differences in the FP system, health care infrastructure or RH climate, providing important direction for policy-makers.