Objective: To assess fertility desires by availability and use of ART/PMTCT services in Rakai, Uganda.

Design: Retrospective analyses of longitudinal data from the Rakai Community Cohort Study.

Methods: Study participants were retrospectively identified and categorized by HIV status. Availability of antiretroviral therapy and prevention of mother-to-child transmission (ART/PMTCT) services in Rakai was defined in three periods: 1) Pre-ART/PMTCT (<2005), 2) ART/PMTCT rollout (2005-06), and 3) universal ART/PMTCT (>2006); and use of ART/PMTCT was coded as yes if the woman received services. Trends in fertility desires were assessed by chi- square. "Modified" Poisson regression was performed using generalized linear models with a log link and Poisson family to estimate prevalence rate ratios (PRR) and 95% confidence intervals of desire for another child among previously and currently pregnant women; PRR were adjusted for demographic and behavioral factors.

Results: 4,227 sexually active women in Rakai, including 436 HIV+ women, contributed 13,970 observations over 5 survey rounds. Fertility desires increased in the population in the ART/PMTCT rollout (adjusted PRR: 1.08, CI: 1.04-1.13) and the universal availability periods (adj. PRR: 1.11, CI: 1.08-1.14) compared to pre-ART/PMTCT period. A total of 862 woman observations used ART/PMTCT services. Fertility desires were similar among ART/PMTCT service users and non-users in cross-sectional analysis (adj. PRR: 0.84, CI: 0.62-1.14) and one year after ART/PMTCT use (adj. PRR: 1.27, CI: 0.83-1.94).

Conclusion: Availability of ART/PMTCT may increase fertility desires of previously pregnant women in Rakai, Uganda. Use of ART/PMTCT services was not correlated with fertility desires of previously or current pregnant women.