Factores explicativos de la adherencia al tratamiento antirretroviral con técnicas de Machine Learning

Objective. To identify explanatory factors associated with adherence to antiretroviral treatment among patients diagnosed with Human Immunodeficiency Virus (HIV).Materials and Methods. A retrospective observational study was conducted on 1,793 patients with confirmed HIV/AIDS diagnosis and antiretroviral treatment indication, affiliated with a Colombian healthcare institution during 2021. Results. The support vector machine model showed 61% predictability in the relationship between antiretroviral treatment adherence and viral load outcome in HIV patients. The variables with the greatest weight in the model, according to the importance ranking, were cardiovascular risk, transmission mechanism, acceptance of the diagnosis, and food consumption. Conclusions. It is essential to recognize that HIV care pathways and models must evolve toward frameworks that integrate governance and social factors. This underscores the need to overcome exclusion, inequity, and barriers that limit timely access to and utilization of healthcare services.