Differences in fertility by HIV serostatus and adjusted HIV prevalence data from an antenatal clinic in northern Uganda

OBJECTIVES: To estimate differences in fertility by HIV serostatus and to validate an adjustment method for estimating the HIV prevalence in the general female population using data from an antenatal clinic. METHODS: We used Cox regression models to retrospectively estimate the age-specific relative fertility (RF) of HIV-positive compared to HIV-negative women among 3314 antenatal clinic attenders in northern Uganda. RF and the age distribution of women in the general female population were used to extrapolate the antenatal clinic-based HIV prevalence. This procedure was indirectly validated by comparing the adjusted estimate with those based on standard adjustment factors derived from general female populations in Uganda and Tanzania. RESULTS: HIV-positive women reported a lower fertility than HIV-negative women [age-adjusted RF = 0.83, 95% confidence interval (CI): 0.75-0.93]. Except for girls aged 15-19 (RF = 0.96, 95% CI: 0.74-1.24) HIV-positive women in all age groups were less fertile (20-24 year: RF = 0.83, 95% CI: 0.67-1.01; 25-29 years: RF = 0.79, 95% CI: 0.62-1.00; 30-49 year: RF = 0.79, 95% CI: 0.65-0.96]. Adjusting the antenatal clinic-based HIV prevalence (11.6%) for these differences yields a higher estimate (13.8%) that is lower than those based on standard adjustment factors derived from general female populations (from 14.6% to 17.7%). CONCLUSIONS: The age-specific pattern of differential fertility by HIV serostatus derived from antenatal clinic data is consistent with findings from population-based studies conducted in Africa. However, differences in fertility between HIV positive and HIV-negative clients underestimate those in the general female population yielding inaccurate estimates when used to extrapolate the HIV prevalence.