Objective: To determine the population effectiveness of a city-wide perinatal HIV prevention program.
Design: An anonymous surveillance of newborn cord blood for HIV serology and nevirapine (NVP).
Methods: All 10 public-sector delivery centers in Lusaka, Zambia participated. All mother-infant pairs delivering during the 12-week surveillance period at the participating centers and who received antenatal care at a public-sector facility in Lusaka were included in the study. The main outcome measure was population NVP coverage, defined as the proportion of HIV-infected women and HIV-exposed infants in the population that ingested NVP.
Results: Of 8787 women in the surveillance population, 7204 (82%) had been offered antenatal HIV testing, of which 5149 (71%) had accepted, and of which 5129 (99%) had received a result. Overall, 2257 of 8787 (26%) were cord seropositive. Of the 1246 (55%) cord blood seropositive women who received an antenatal HIV test result, 1112 (89%) received a positive result; the other 134 comprise seroconverters and clerical errors. Only 751 of 1112 (68%) women who received a positive antenatal test result and a NVP tablet for ingestion at labor onset had NVP detected in the cord blood (i.e., maternal non-adherence rate was 32%). A total of 675 infants born to 751 adherent mothers (90%) received NVP before discharge. Thus, only 675 of 2257 (30%) seropositive mother-infant pairs in the surveillance population received both a maternal and infant dose of NVP.
Conclusions: Successful perinatal HIV prevention requires each mother-infant pair to negotiate a cascade of events that begins with offering HIV testing and continues through adherence to the prescribed regimen. This novel surveillance demonstrates that failures occur at each step, resulting in reduced coverage and diminished program effectiveness.