Sentinel surveillance for HIV-1 among pregnant women in a developing country: 3 years' experience and comparison with a population serosurvey.

Between October 1989 and September 1991 health workers took blood samples from pregnant women attending Makongoro Clinic in Mwanza Tanzania to determine age-specific HIV-1 seroprevalence in pregnant women trends over 3 years and the feasibility sustainability and costs of HIV monitoring and to compare age-specific seroprevalences in pregnant women with those of all women. Overall HIV prevalence among pregnant women stood at 11.5%. Even though the younger and older age groups had lower seroprevalence than the other age groups (10% for those under 20 years old 11% for 30-34 year olds and 8.3% for = or > 35 year olds vs. 12.8% for 25-29 year olds and 12.3% for 20-24 year olds) the differences were not significant. HIV seroprevalence appeared to increase in 1988-89 in most age groups but no significant evidence of a linear trend occurred during the study period for any age group. This absence of significant change in HIV seroprevalence over 3 years likely indicated a considerable HIV incidence. The population serosurvey revealed an HIV seroprevalence of 15.1% among the general adult female population suggesting that the sentinel surveillance among pregnant women could have greatly underestimated population HIV seroprevalence for women under 35 years old (p = .02). Sentinel surveillance improved the clinics ability to detect anemia and reintroduced syphilis screening. Unlinked anonymous testing resulted in additional recurrent costs of S$1.7/specimen. Syphilis screening added recurrent costs of US$0.57/woman. These findings indicated that health workers in Africa can successfully introduce and maintain anonymous HIV-1 sentinel surveillance of pregnant women and this can be accomplished with concurrent syphilis screening.