Human immunodeficiency virus infection and infective endocarditis among injecting drug users.

Human immunodeficiency virus (HIV) infection and infective endocarditis are serious complications of injection drug use. To determine whether HIV infection may increase the risk of endocarditis beyond that associated with drug injection, we performed a nested case-control study among injecting drug users taking part in an ongoing cohort. We identified 26 participants with infective endocarditis between cohort enrollment (in 1988-1989) and June 1992, through reviews of medical records and death certificates. We matched each endocarditis case with up to five controls (N = 120) on enrollment date, race/ethnicity, and follow-up time. Data were taken from baseline and from the one follow-up visit: the last visit before the endocarditis occurred for cases and the closest visit (+/- 3 months) for controls. We used conditional logistic regression to quantify the association between HIV serostatus at follow-up and subsequent endocarditis, after adjusting for a history of endocarditis or sepsis before enrollment, injection duration, current injection frequency, and a recent history of abscess at injection sites. Among current injectors at follow-up, the adjusted odds ratio (OR) of developing endocarditis for HIV-seropositive subjects with > or = 350 CD4 cells per microliter, compared with HIV-seronegative subjects, was 2.31 [95% confidence interval (CI) = 0.61-8.78]; the corresponding OR for HIV-seropositive subjects with < 350 CD4 cells per microliter was 8.31 (95% CI = 1.23-56.37). These data indicate that HIV-related immunodeficiency may independently increase the risk of infective endocarditis among injecting drug users.

The effect of latent Mycobacterium tuberculosis infection on human immunodeficiency virus (HIV) disease progression and HIV RNA load among injecting drug users.

To examine the relationship between latent Mycobacterium tuberculosis infection and human immunodeficiency virus (HIV) disease progression, two studies were done among a cohort of HIV-infected injecting drug users. First, the decline in CD4 cell count after baseline tuberculin skin testing was prospectively compared for 37 tuberculin-positive (induration > or = 5 mm) and 284 tuberculin-negative (induration < or = 2 mm) persons. After adjustment for baseline immune function, the mean 6-month CD4 cell decline was not significantly different (34.5 vs. 45.6 cells, respectively, P = .14). Second, the plasma HIV burden at baseline skin testing was compared for 33 tuberculin-positive cases and 33 matched tuberculin-negative controls. HIV RNA was detected in 8 cases and 10 controls (odds ratio = 0.67, 95% confidence interval = 0.19-2.36). Among the 14 pairs with HIV detected in > or = 1 member, the HIV concentration was higher for the case in 4 and for the control in 10 (P = .18). These findings suggest that unlike active tuberculosis, latent M. tuberculosis infection does not hasten HIV progression.