| Sumario: | Human immunodeficiency virus (HIV) remains a public health issue in the U.S., affecting approximately 1.2 million individuals, many of whom are unaware of their infection status. This study reviews predictors and the performance of HIV risk prediction models. We analyzed 18 studies published since 2010, which featured logistic regression, survival analysis, and machine learning techniques. These studies focused on diverse populations, including men who have sex with men, emergency department visitors, and the general population. Key predictors of HIV risk included demographics (age, sex, race) and behavioral factors (sexual practices, drug use). Electronic health records (EHR) documenting diagnoses of sexually transmitted infection (STI) were significant in all models. Behaviors like condomless sex, multiple sexual partners, and drug use were also strongly linked to increased risk scores. However, we noted a lack of social determinants of health in risk models, and a gap in studies focusing on cis female and transgender populations.
|
|---|