|
Predicting pregnancy among HIV-1-discordant couples
B.L. Guthrie1, A. Gatuguta2, R. Bosire3, R. Choi1, R.D. Mackelprang1, J. Kiarie2,4, G. John-Stewart1, C. Farquhar1
1University of Washington, Seattle, United States, 2University of Nairobi, Nairobi, Kenya, 3Kenya Medical Research Institute (KEMRI), Nairobi, Kenya, 4Kenyatta National Hospital (KNH), Nairobi, Kenya
Background: The number of stable HIV-1-discordant couples is increasing and desire for children among these couples is common. Increasingly, intervention trials are focusing on discordant couples, making definition of incidence and predictors of pregnancy important for designing rational HIV-1 prevention strategies in discordant couples.
Methods: HIV-1-discordant couples from Nairobi, Kenya were followed for up to 2 years and assessed quarterly for pregnancies. Cox regression was used to assess predictors of pregnancy.
Results: 275 women in discordant relationships were followed: 183 (66.5%) were HIV-1-infected and 92 (33.5%) were uninfected. One-year cumulative incidence of pregnancy was 11.1% (95%CI, 7.6-16.0%). Significant predictors of pregnancy at the time of enrollment were self-reported desire to have children in the future (HR=5.56; 95%CI, 2.26-13.67) and age (HR=0.91; 95%CI, 0.86-0.98). The majority of pregnancies occurred among women < 30 years old who reported a desire for future children at enrollment (1-year incidence 26.2%). Women ≥30 years old who reported a desire for future children had an intermediate pregnancy rate (10.3%), while women < 30 years and ≥30 years of age who did not report a desire for future children had low pregnancy rates (1.5% and 2.3%, respectively). Pregnancy rates did not differ between HIV-1-infected and uninfected women (HR=0.90; 95% CI, 0.42-1.94).
Conclusions: We found a high pregnancy rate among women in HIV-1-discordant couples, indicating desire for children may override concerns of transmission and increase unprotected sex. These findings emphasize the importance of addressing pregnancy when planning and conducting prospective studies of discordant couples. Excluding couples most likely to become pregnant may simplify follow-up, but can significantly impact generalizability. Pregnancies are inevitable in prospective cohorts, and researchers should anticipate pregnancies and pregnancy desire. At a broader level, these findings highlight the need for interventions such as pre-exposure prophylaxis and microbicides to make conception safer for discordant couples.
|