The prevalence of human immunodeficiency virus (HIV) in correctional facili
ties is much higher than in the general population. However, HIV prevention
resources are limited, making it important to evaluate different preventio
n programs in prison settings. Our study presents the cost-effectiveness of
offering HIV counseling and testing (CTI to soon-to-be-released inmates in
US prisons. A decision model was used to estimate the costs and benefits (
averted HIV cases) of HIV testing and counseling compared to no CT from a s
ocietal perspective. Model parameters were HIV prevalence among otherwise u
ntested inmates (1%); acceptance of CT (50%); risk for HIV transmission fro
m infected individuals (7%); risk of HIV acquisition for uninfected individ
uals (0.3%); and reduction of risk after counseling for those infected (25%
) and uninfected (20%). Marginal costs of testing and counseling per person
were used Ino fixed costs). If infected, the cost was $78.17; if uninfecte
d, it runs $24.63. A lifetime treatment cost of $186,900 was used to estima
te the benefits of Prevented HIV infections. Sensitivity and threshold anal
ysis were done to test the robustness of these parameters. Our baseline mod
el shows that, compared to no CT, offering CT to 10,000 inmates detects 50
new or previously undiagnosed infections and averts 4 future cases of HIV a
t a cost of $125,000 to prison systems. However, this will save society ove
r $550,000. Increase in HIV prevalence, risk of transmission, or effectiven
ess of counseling increased societal savings. As prevalence increases, focu
sing on HIV-infected inmates prevents additional future infections; however
, when HIV prevalence is less than 5%, testing and counseling of both infec
ted and uninfected inmates are important for HIV prevention.