One in seven people living with HIV (PLH) in the U.S. cycle throughout the criminal justice system each year. While studies show that up to 70% of inmates living with HIV who are prescribed ART can achieve viral suppression during their incarceration, this benefit is not sustained after release into the community. In order to help decrease HIV-related morbidity and mortality among former inmates with HIV and to reduce HIV transmission to their community contacts, more information is needed regarding longitudinal HIV outcomes as inmates cycle between the criminal justice system and the community. Connecticut (CT) is unusual in that it has an integrated correctional system which, combined with CT's statewide HIV/AIDS surveillance monitoring system (eHARS), provides a novel opportunity to study the longitudinal interactions between HIV continuity of care and criminal justice involvement. By linking multiple CT Department of Correction databases with CT's HIV/AIDS surveillance system, this F30 application is poised to determine the relationship between incarceration, release into the community, and the continuity of HIV care. By using a retrospective cohort of PLH released from CT prisons and jails between 2007-2013, the Specific Aims for this application are to: (1) use survival analysis to assess HIV treatment outcomes and engagement in care after release from correctional facilities in CT; (2) use time-varying effect modeling to evaluate longitudinal trends in HIV viral suppression during incarceration compared to time in the community, based on type and extent of interaction with the justice system; and (3) use group-based trajectory mixture modeling to identify groups of individuals demonstrating distinct viral load trends indicative of poor HIV management. Exploratory aims are to identify and describe HIV outcomes for particularly high-risk (i.e., female and ethnic/racial minority) sub-populations through both hypothesis-drive analysis and data mining. The innovative linkage and analysis of community-based clinical data and integrated correctional data at a statewide level will allow for the first study to follow both recidivists and non-recidivists in the community and o use advanced longitudinal methods to evaluate HIV outcomes during time in the community as compared to imprisonment. The proposed study will identify key time points at which former inmates are at highest risk for developing poor HIV outcomes. Findings will help community and correctional system-based healthcare interventions target the highest risk groups of individuals at the time points that would be most beneficial. Through this three-year fellowship, the applicant will gain training in advanced biostatistical methods for the analysis of longitudinal an big data with mentorship from a multidisciplinary research team at Yale University. The training plan outlined in this application aims to prepare her to apply for an NIH K/R award, furthering her career as an independent clinician-investigator generating data to directly inform health policy and clinical interventions to effectively treat HIV/AIDS in vulnerable populations.
Research suggests that people living with HIV (PLH) who are involved in the criminal justice system experience suboptimal HIV treatment outcomes after release from jail or prison into the community. The proposed study will integrate correctional data with HIV/AIDS surveillance data to follow PLH released from CT jails or prisons during 2007-2013, in order to: (1) demonstrate applying 'big data' to identify PLH who have fallen out of care; (2) use advanced longitudinal analytical methods to identify policy-relevant points in time at which PLH involved in the criminal justice system are most vulnerable to poor HIV treatment outcomes after release into the community; and (3) use advanced modeling methods to identify populations of former inmates who could most benefit from specific interventions to improve HIV treatment outcomes post-release. Findings will inform correctional and community-based health policies by helping to target individuals at greatest risk for poor HIV treatment outcomes and identify crucial time points at which these interventions would be most beneficial.
