HIV retesting during pregnancy/postpartum is crucial for early detection and treatment of incident maternal HIV infection, and to achieve elimination of mother-to-child HIV transmission (MTCT). Kenyan guidelines recommend retesting peripartum HIV negative women but data on implementation are lacking. We measured the frequency of HIV retesting during pregnancy, delivery, and postpartum and correlates of postpartum retesting.
HIV-seronegative women seeking maternal and child health (MCH) services were enrolled in a cross-sectional study in rural Kenya at the Ahero County and Bondo sub-County Hospitals at one of the following time points: pregnancy; delivery; 6 weeks, 6 months, or 9 months postpartum. Data on programmatic retesting was abstracted from MCH booklets to ascertain retesting during pregnancy and/or postpartum prior to the study visit. Retesting was defined as any HIV test after the initial antenatal care (ANC) test or after pregnancy if testing was not done in ANC. Poisson regression, clustered by site, was used to identify correlates of retesting among women enrolled at 9 months postpartum.
Among 1919 women enrolled, the median age was 23 years, 63% were married and the median number of times tested for HIV in the most recent pregnancy/postpartum period was 1 (interquartile range [IQR]: 1-2). Overall, 659 women were enrolled in the 3rd trimester, 128 within 48 hours after delivery, 387 at 6 weeks postpartum, 412 at 6 months postpartum, and 333 at 9 months postpartum. Prevalence of any programmatic HIV retesting was significantly higher at 6 weeks postpartum (46%) than in the 3rd trimester (23%), at delivery (5%), and at 6 months postpartum (28%) (p<0.001 for all). By 9 months postpartum, HIV retesting was associated with prior sexually transmitted infection (STI) diagnosis (Prevalence Ratio [PR]:1.28, 95% Confidence Interval [CI]:1.06-1.56; p<.001), higher gravidity (PR:1.05 per pregnancy, 95% CI:1.04-1.06; p<.001), and being an orphan (PR:1.02, 95% CI:1.01-1.02 p=.02). Results were similar in a multivariable analysis of cofactors significant in the univariate model.
Prevalence of retesting was higher in the early postpartum period and more common among women who had a history of STIs and higher gravidity. Strategies to offer retesting to all peripartum women in high prevalence regions could help identify incident maternal HIV and maximize prevention of MTCT efforts.