A comparative analysis of three methods for detecting potential signals of HIV acquisition using recent infection surveillance data from Malawi in 2023

Abstract Recent infection surveillance aims to classify newly diagnosed HIV infections as recent (acquired approximately ≤12 months ago) or long-term to identify geographic areas and subpopulations with potentially elevated HIV acquisition. As this surveillance evolves, programs may evaluate different methods for detecting potential signals of acquisition. Three methods were used in Malawi to analyze recent infection surveillance data from 289 health facilities. The first method utilized a numeric threshold of recent infections per month disaggregated by facility-level classification; the second method used increases in recent infections over a 6-month period ≥2 standard deviations above the mean for the previous 12 months for a demographic group or geography; and the third method utilized a spatiotemporal threshold of expected recent infections among an at-risk population using a Poisson probability model at 6-month intervals. In total, 89 unique facilities and 1 district were identified as having potentially elevated HIV acquisition, with 29 appearing in multiple methods. Using more than one method of identifying potential signals of acquisition may help prioritize facilities for data triangulation and public health response activities. Methods of identifying and responding to ongoing HIV acquisition should ultimately be based on available data sources, the country’s epidemic control goals, feasibility, acceptability, and sustainability.