Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: Rates of both HIV and mental illness (MI) are high in prison populations. Little is known about characteristics of HIV+ inmates with MI and their post-release needs, though research suggests greater recidivism among MI HIV+ offenders (Harris, Raffi, Tonge, Uldall, 2002; Marushak, 2001). Using interviews of HIV+ inmates, the aims of this study were to: 1) determine the proportion of MI in a HIV+ prison population; 2) test for differences between MI and non-MI HIV+ inmates; 3) determine if MI is significantly associated with higher levels of community need, poorer health, lower levels of coping self-efficacy; 4) consider implications of findings for practice.

Methods: This investigation analyzed responses from quantitative interviews of 95 HIV+ North Carolina inmates within three months of prison release. Proportion of MI was determined using a composite variable from chart data, individual report, and the CES-D. Bivariate analyses examined differences between MI and non-MI inmates. Logistic regression analysis was used to determine if MI was significantly associated with having more unmet needs in the six months pre-incarceration, more anticipated post-release needs, poorer HIV health status, and lower coping self-efficacy.

Results: Chart review and individual report indicated 34(35.8%) in the sample had a MI. Including individuals with clinically significant CES-D scores, an additional 23 (24%) were identified as MI. Bivariate analysis showed MI inmates were significantly less likely to be Black (73% vs. 97%, p<.01), complete high school (47% vs. 81%, p<.01) , or work in the six months pre-incarceration (39% vs. 60%, p<.05); no differences were seen in age, gender, CD4 count above 200, or detectable viral load. Bivariate analysis also showed significant differences between MI and non-MI in unmet need variables and coping self-efficacy; these were next examined using multivariate analysis. Logistic regression (controlling for age, race, gender, drug use) indicated MI was significantly associated with going without healthcare to meet basic needs (O.R.=4.00, 95%CI 1.64-9.76), needing housing (O.R.=2.88, 95%CI 1.19-6.99), and having a place for HIV care pre-incarceration (O.R.=4.00, 95%CI 0.12-0.97). MI was significantly associated with anticipated need for help with housing (O.R.=4.52, 95%CI 1.66-12.30), drug treatment (O.R.=2.45, 95%CI 1.03-5.81), alcohol treatment (O.R.=2.90, 95%CI 1.11-7.54), transportation (O.R.=5.42, 95%CI 1.91-15.38), medication adherence (O.R.=5.00, 95%CI 1.51-16.6), community re-integration post-release (O.R.=2.66, 95%CI 1.07-6.66), and lowered coping self-efficacy (O.R.=0.21, 95%CI 0.08-0.54).

Practice Implications: Findings indicate up to two-thirds of HIV+ inmates may have MI, with 40% of these individuals unidentified. Social workers in correctional settings should advocate for regular MI screening of HIV+ inmates, including during release planning. Results also suggest that HIV+ inmates with MI experienced more unmet needs pre-incarceration, have more psychosocial needs upon release, and have lower ability to cope with these challenges. Release planning for HIV+ inmates with MI requires more than focus on HIV medical care and should consider the elevated community needs and lowered self-efficacy of this population.

Harris, V., Raffi, R., Tonge, S., Uldall, K. (2002). Rearrest: Does HIV serostatus make a difference? AIDS Care, 14: 839-849. Marushak, L. (2001). HIV in Prisons, 1999. Bureau of Justice Statistics Bulletin. NCJ187456.