Search

Introduction Setting and monitoring progress towards targets for HIV control is critical in ensuring responsive programmes. Here, we explore how to apply targets for reduction in HIV incidence to local settings and which indicators give the strongest signal of a change in incidence in the population and are therefore most important to monitor. Methods We use location‐specific HIV transmission models, tailored to the epidemics in the counties and major cities in Kenya, to project a wide range of plausible future epidemic trajectories through varying behaviours, treatment coverage and prevention interventions. We look at the change in incidence across modelled scenarios in each location between 2015 and 2030 to inform local target setting. We also simulate the measurement of a library of potential indicators and assess which are most strongly associated with a change in incidence. Results Considerable variation was observed in the trajectory of the local epidemics under the plausible scenarios defined (only 10 of 48 locations saw a median reduction in incidence of greater than or equal to an 80% target by 2030). Indicators that provide strong signals in certain epidemic types may not perform consistently well in settings with different epidemiological features. Predicting changes in incidence is more challenging in advanced generalized epidemics compared to concentrated epidemics where changes in high‐risk sub‐populations track more closely to the population as a whole. Many indicators demonstrate only limited association with incidence (such as “condom use” or “pre‐exposure prophylaxis coverage”). This is because many other factors (low effectiveness, impact of other interventions, countervailing changes in risk behaviours, etc.) can confound the relationship between interventions and their ultimate long‐term impact, especially for an intervention with low expected coverage. The population prevalence of viral suppression shows the most consistent associations with long‐term changes in incidence even in the largest generalized epidemics. Conclusions Target setting should be appropriate for the local epidemic and what can feasibly be achieved. There is no one universally reliable indicator to predict future HIV incidence across settings. Thus, the signature of epidemic control must contain indications of success across a wide range of interventions and outcomes.

Objective: There is evidence of substantial subnational variation in the HIV epidemic. However, robust spatial HIV data are often only available at high levels of geographic aggregation and not at the finer resolution needed for decision making. Therefore, spatial analysis methods that leverage available data to provide local estimates of HIV prevalence may be useful. Such methods exist but have not been formally compared when applied to HIV. Design/methods: Six candidate methods – including those used by the Joint United Nations Programme on HIV/AIDS to generate maps and a Bayesian geostatistical approach applied to other diseases – were used to generate maps and subnational estimates of HIV prevalence across three countries using cluster level data from household surveys. Two approaches were used to assess the accuracy of predictions: internal validation, whereby a proportion of input data is held back (test dataset) to challenge predictions; and comparison with location-specific data from household surveys in earlier years. Results: Each of the methods can generate usefully accurate predictions of prevalence at unsampled locations, with the magnitude of the error in predictions similar across approaches. However, the Bayesian geostatistical approach consistently gave marginally the strongest statistical performance across countries and validation procedures. Conclusions: Available methods may be able to furnish estimates of HIV prevalence at finer spatial scales than the data currently allow. The subnational variation revealed can be integrated into planning to ensure responsiveness to the spatial features of the epidemic. The Bayesian geostatistical approach is a promising strategy for integrating HIV data to generate robust local estimates.