Evaluating two adjustment methods to extrapolate HIV prevalence from pregnant women to the general female population in sub-Saharan Africa

Objective: To evaluate two methods for estimating HIV prevalence among the general female population of reproductive age by adjusting data observed among antenatal clinic (ANC) attendees. Methods: We adjusted the HIV prevalence among ANC attendees in Fort Portal (Uganda; 1994–1995), Mwanza municipality (Tanzania; 1990–1991), rural Mwanza (Tanzania; 1991–1993), Mposhi district (Zambia; 1994), Chelston (Lusaka, Zambia; 1994, 1996 and 1998) and Ndola (Zambia; 1998), using firstly a method that accounts for differences in age-specific fertility by HIV serostatus and secondly a method that accounts for differences in HIV prevalence by fertility risk category and parity. Results: The non-adjusted HIV prevalence among ANC attendees underestimates the prevalence among the general female population by 8.0% in Chelston in 1998 and by between 20.7% and 31.9% in all other cases. The adjusted prevalence obtained using the first method underestimates the prevalence among the general female population by about 0.5% in Fort Portal and Mposhi; it overestimates that observed in Chelston in 1994 and 1996 by about 3.5%, and that observed in Ndola, urban Mwanza and rural Mwanza, by 6.5%, 10.6% and 12.8%, respectively. The second method (applied for only four sites) provides an overestimate of 7.0% in Chelston in 1994 and an underestimate of 3.8% and 2.1% in Ndola and rural Mwanza, respectively. Both adjustment methods overestimate the 1998 prevalence in Chelston, producing less accurate estimates than the non-adjusted data. Conclusions: The HIV prevalence among women in the general population could be estimated fairly accurately by these methods in settings with mature epidemics.