Back

The importance of adolescent girls and \"epidemic gearing\" on HIV prevalence across West Africa

Prudden, H.; Mukandavire, Z.; Gorgens, M.; Wilson, D.; Panovska-Griffiths, J.; Watts, C.

2019-10-18 hiv aids
10.1101/19008839 medRxiv
Show abstract

BackgroundIn West Africa HIV prevalence varies between 0.1-6% in female and between 0.1-4% in the male general population. Male circumcision is almost universal, and it is unclear what drives this variation. We use mathematical modelling to identify the determinants of this variation across fourteen West African countries. MethodsWe developed a novel dynamic model of HIV transmission between population cohorts of female sex workers (FSWs), their clients, females with 2+ partners in the past year and other sexually active women and men in the general population. Parameter ranges were determined from the literature and sampled using Latin Hypercube sampling to identify parameter sets that fit West African HIV prevalence data. Partial-rank correlation coefficients between different model parameters and the HIV prevalence in general male and female population across 14 countries were calculated to determine to most significantly correlated model parameters to HIV prevalence. ResultsThe key determinant of HIV in females when prevalence is between 0-3% is the size of the brothel and non-brothel FSW groups. When female HIV prevalence >3%, the percentage of sexually active adolescent females with 2+ partners has greater influence on HIV prevalence. The size of the FSW groups has the most significant impact on HIV prevalence for males. ConclusionsOur findings confirm the role of FSWs in West Africa as an important determinant of HIV risk, but also identify, in countries with higher HIV prevalence, the emerging role of a group of adolescent girls with 2+ partners is an important determinant of risk. In fact, our findings suggest that this group may enable the epidemic to be effectively "geared up" when partnerships are formed with higher-risk males, indicating additional prevention needs amongst this group. FundingThis study was funded by UNAIDS.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.