Objective: We conducted a systematic review of HIV progression models to identify the mathematical structures used, the main research questions and key model aspects in terms of quality and robustness.
Methods: We searched for articles published before February 2009 that described models of HIV progression in humans. We included two strategies of search with and without MeSH terms. We classified the models by their mathematical structure and research question. We created a checklist of desirable features of the models, reviewed and classified the articles to inform our conclusions.
Results: Among 3491 articles found, 93 met the inclusion criteria. Among the selected articles, 60 used transition models, 25 applied differential equations, and eight had other structures. We did not find a relation between the type of question explored and the modeling method used. None of the studies complied with the complete set of items in the checklist, but 6.5% cover at least 90% of them.
Conclusion: There is an enormous heterogeneity of HIV modeling exercises in terms of methods used and topics addressed, as well as in the presentation of key aspects of the articles in terms of quality and robustness.