Introduction: Given the current challenges to statistical backcalculation techniques, alternative approaches are being pursued for modeling the HIV/AIDS epidemic within subpopulations. The objective of this study was to estimate HIV incidence and detection rates among specific subgroups in Louisiana using a flexible staged Markov model approach based on that developed by Aalen and colleagues (Stat Med,1997).
Methods: Time-inhomogeneous discrete Markov models of disease progression, detection, and treatment were applied to quarterly Louisiana HIV/AIDS surveillance data, including HIV (non-AIDS) data, from 1981 to 1996. Estimated parameters included stage-specific probabilities of detection and non-parametric estimates of incidence. Fixed parameters described disease progression, treatment uptake, and treatment effect. Parameters were permitted to vary over time. The model accounted for reporting delays, pre-AIDS mortality, the 1993 change in AIDS case definition, and missing risk information.
Results: Incidence and detection probabilities were modeled for MSM, IDUs, high risk heterosexuals, white men, African-American men, and women. Recent estimates of incidence, prevalence, and stage-specific magnitudes of undetected and untreated infections were provided for each subgroup. Relative comparisons of estimated parameters and results were particularly effective for understanding differences between subgroups.
Implications: Although subject to some of the issues regarding backprojection techniques, the approach meets a data need for local program planning, particularly with respect to projections of inadequate treatment uptake and/or response between subgroups. From a surveillance standpoint, the results contribute to an enhanced understanding of how the interpretation of HIV surveillance trends may be biased by differential detection rates between subgroups.
Learning Objectives: N/A
Keywords: Surveillance,
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.