550 
Characteristics and Behaviors Associated with HIV Infection in a Large Southern Prison System
David Rosen*1,2, D Wohl1,2, B White2, and V Schoenbach1
1Univ of North Carolina at Chapel Hill Sch of Publ Hlth, US and 2Univ of North Carolina at Chapel Hill Sch of Med, US
Background: The prevalence of HIV among correctional
populations is many times that of the general U.S. population. Inmates in 28 state
prison systems are provided voluntary (i.e. consent-based) HIV testing. The
purpose of this study was to characterize the population of HIV+
prisoners in a state prison system with voluntary HIV testing and to estimate
the number of undetected HIV cases.
Methods: Using de-identified electronic records of adult
prisoners entering the North Carolina Department of Corrections from January
2004 to May 2006, associations between HIV serostatus and conventional HIV risk
behaviors, mental health, co-infection status, and sociodemographic
characteristics were estimated using logistic regression. The range of
undetected cases was estimated by multiplying the age-, sex-, and race-specific
number of untested prisoners by corresponding stratum-specific HIV prevalence
from both tested prisoners and from data for the entire state.
Results: Approximately 40% (n = 21,419) of
prisoners were tested for HIV, and nearly 3.4% (n = 718) of these tested
prisoners were HIV+. Among men, HIV infection was most strongly
associated with men who have sex with men (MSM) (OR 8.0), black race (OR 6.2),
other non-white race (OR 7.4), and age 35 to 44 years (OR 4.1). The strongest
risk factor among women was black race (OR 3.8). Modest associations (i.e. OR
<3.0) were observed for several other risk factors across employment,
education, mental health, imprisonment, drug use, and co-infection domains. Among
HIV+ prisoners, 65% were co-infected with hepatitis C virus (HCV),
>80% of HIV+ prisoners had been initially diagnosed with HIV
before entering prison, 30% had “ever been told you have AIDS,” but fewer than
half reported a conventional HIV risk behavior. We estimated between 23% (217)
and 67% (1478) of HIV cases remained undetected.
Conclusions: Modest to moderate associations and the
large proportion of prisoners without reported risks highlight the potential
limitations of risk factor-based HIV testing in prison. Our most conservative
estimate suggests that about one-fourth of all HIV cases in the North Carolina
Department of Corrections remain undetected.
|