Primary Submission Category: Mortality
Impact of decarceration, alcohol, substance use, and mental health screening on life expectancies of Black sexual minority men and Black transgender women (BSMM/BTW) living with HIV in the United States: A Simulation Study based on HPTN 061
Authors: Jonathan Feelemyer Scott Braithwaite Joy Scheidell Russell Brewer Typhanye Dyer Charles Cleland Christopher Hucks-Ortiz Amy Justice Kenneth Mayer Ames Grawert Jay Kaufman Anna Bershteyn Maria Khan
Presenting Author: Maria Khan*
Background
Black sexual minority men and Black transgender women with HIV (BSMTW-HIV) experience high rates of incarceration, substance use, overdose, and lower life expectancies (LE) than the general US population. We estimated the impact of decarceration (i.e., the effort to limit the number of people who are detained) and screening for psychiatric conditions and substance use on life expectancy of US BSMTW-HIV using a microsimulation model of the HIV Prevention Trials (HPTN) 061, which enrolled BSMTW from six US cities.
Methods
We augmented a microsimulation model previously validated to predict life expectancy and leading causes of death in the US with estimates from HPTN 061 for how recent incarceration (past six months) independently predicts subsequent binge drinking (RR=1.14), stimulant use (RR=1.53), tobacco use (RR=1.77), depression (RR=1.09), and anxiety (RR=1.09). Data from another large US HIV cohort, the Veterans Aging Cohort Study, were used to estimate independent associations among psychiatric and substance use disorders, such that we were able to simulate the influence of treatment of one condition on improvement on others (e.g., treating depression reduces opioid use). The simulation also incorporated relationships gleaned from systematic reviews regarding how substance use and depression, anxiety, and pain lead to non-adherence to treatment for HIV and other conditions. We used this augmented simulation to estimate LE for BSMTW-HIV with a history of incarceration under alternative policies of decarceration (i.e., reducing the fraction of the cohort exposed to incarceration), screening for psychiatric conditions and substance use, or both.
Results
At baseline, mean age of this microsimulation analytic dataset was 39.9 years old (median: 42 years, IQR: 31-48). LE (at birth) for BSMTW-HIV was 61.3 years (95% confidence interval (CI): 61.2-61.4) without reductions in incarceration or improving screening and treatment. Reducing incarceration by 25%, 33%, 50%, and 100% increased LE by 0.29 years, 0.31 years, 0.53 years, and 1.08 years, respectively, versus no reductions in incarceration. When reducing incarceration by 33%, a feasible decarceration target, and implementing screening for alcohol, tobacco, substance use, and depression, in which a positive screen triggers diagnostic assessment for all psychiatric and substance use conditions and linkage to treatment, LE increased by 1.52 years compared to no screening or decarceration.
Discussion/Interpretation
LE among BSMTW-HIV is short compared with other people with HIV (range depending on risk factors and antiretroviral treatment initiation and adherence: 70-80 years). Reducing incarceration and improving screening and treatment of psychiatric conditions and substance use could substantially increase LE in this population, but further research is needed to identify other factors contributing to the lower LE.