Interpreting neonatal HIV seroprevalence data in Great Britain: the importance of differential fertility

In order to derive more accurate estimates of HIV infection among the general population in Great Britain from unlinked anonymous neonatal seroprevalence data, the differential HIV and fertility patterns amongst groups of women were explored. The HIV risk in women was estimated using neonatal seroprevalence data and reports of diagnoses of HIV infection in pregnant women. Live birth rates were estimated using population data and these were supplemented using data from a prospective European cohort of HIV-infected pregnant women. HIV prevalence was higher in women born in sub-Saharan Africa (SSA) and injecting drug users (IDUs) (2.50% and 0.29% respectively in London) compared to other women (0.068%). Fertility was also higher in women born in SSA (OR 1.33 using population data), whilst IDUs were more likely to have had a previous termination (OR 1.48 using the European cohort of HIV-infected pregnant women data). We conclude that when unlinked anonymous neonatal seroprevalence data is used to estimate general population prevalence, adjustments need to be made for fertility differentials in population subgroups at varying risk of HIV.