The purpose of this study is to determine if associations exist between the contexts in which people obtain healthcare (i.e., healthcare context) or live (i.e., residential context) and each of five outcomes (HIV testing, receipt of tes results, linkage to HIV/AIDS care, retention in HIV care and HIV viral load). The epidemiology of HIV/AIDS among racial/ethnic minorities (specifically, Blacks and Latinos) and older adults (i.e., 50 and older) suggests they may encounter barriers that contribute to disparities in early detection of HIV and in their prognoses. Hence, we will also examine disparities in the outcomes by race/ethnicity and older age (age >50). Drawing on the socio-behavioral sciences, the study will inform clinical practice in ways that promote equity in care across diverse groups, neighborhood conditions and stages of adulthood. The study's primary Specific Aims are to: (1) Examine relations between healthcare context, residential context and HIV testing during primary care visits using logistic regression multilevel models with random effects based on primary care patients'electronic medical records;and, (2) Examine relations between healthcare context, residential context and receipt of HIV test results and linkage to HIV/AIDS care, respectively, among managed care enrollees newly diagnosed as HIV-positive using multilevel logistic regression and Cox proportional hazards models with random effects. The secondary Specific Aim is to determine if racial/ethnic- or age-related (i.e., aged <50 vs. >50 years) disparities exist in these relationships. Building on our work on HIV testing and care, and guided by a model integrating the Public Health Critical Race Praxis and Behavioral Model of Healthcare Utilization, the four-year study based on the electronic medical records (EMRs) of adults enrolled in the largest managed care organization in the region. We will pool data over five years (2007-2011) to examine HIV testing among all patients presenting for primary care, and to examine receipt of HIV test results, linkage to and retention in HIV/AIDS care as well as HIV viral load among all patients newly diagnosed as HIV-positive. The study will comprise four multilevel analyses of patients'residential contexts (e.g., neighborhood HIV prevalence) and healthcare contexts (e.g., characteristics of the patient population) relative to HIV testing durin primary care visits (Aim 1), receipt of HIV test results and linkage to HIV/AIDS care among those diagnosed as HIV-positive (Aim 2), retention in HIV/AIDS care and HIV RNA viral load up to one-year post diagnosis. We will use personal and geospatial codes in the EMRs to link to: (1) files containing detailed information on each provider (e.g., demographics, specialty);(2) public data from the Centers for Disease Control and Prevention on HIV prevalence and HIV test sites in each zip code;(3) 2010 Census socioeconomic data (e.g., concentrated poverty) for each zip code;and, (4) global positioning system software to calculate the distance from a patient's home to their provider. This interdisciplinary, inter- institutional collaboration leveraes the expertise of a diverse team of new and seasoned investigators.
The goal of the proposed research is to learn why adults, especially racial/ethnic minorities and older adults, who have insurance may nevertheless not receive HIV testing or HIV/AIDS care at the recommended levels. The study compares MCO enrollees based on their medical records, assigned providers and neighborhood social conditions to see if certain patients have a harder time getting testing or care because of the type of provider they see or because of where they live.
|Ford, Chandra L; Mesfin, Mulatu S; Gaines, Tommi L et al. (2016) How Underestimates of Need Contribute to Biased Conclusions. Sex Transm Dis 43:696-697|