[Progress in research of estimation methods for HIV infection incidence]

HIV infection incidence reflects the level of epidemic activity and temporal trends of the HIV/AIDS epidemic, which can be used as a key indicator for the evaluation of prevention and control effectiveness to inform intervention strategies. However, because the exact time of HIV infection is difficult to determine for most individuals living with HIV, the incidence cannot be directly measured, and its estimation is a central topic in HIV epidemiological research. At present, the major approaches for estimating HIV infection incidence are as follows. The first is prospective cohort study, which follows HIV-negative individuals over time and directly observes seroconversion events to calculate incidence, and is widely regarded as the gold standard. The second approach is laboratory-based recent infection testing, which identifies recent infections based on dynamic immunological and virological changes following infection and is subsequently used to estimate incidence at population level. The third is continuous testing data analysis, which uses routine clinical testing records and repeated cross-sectional surveillance data to identify repeated tests and indirectly estimate incidence. The fourth is mathematical modelling which integrates surveillance data, demographic information, and epidemiological parameters to reconstruct incidence level and epidemic trajectory. Each of the four approaches is suitable to different research settings. They vary in data requirements, implementation costs, and estimation accuracy, with distinct strengths and limitations. This paper summarizes the principles, applications, and limitations of four major approaches, i.e. prospective cohort study, laboratory-based recent infection testing, continuous testing data analysis, and mathematical modelling, to provide methodological reference to understand HIV infection dynamics and estimate the incidence.