AIDS and injecting drug use: survival determinants in the highly active antiretroviral therapy era

Objectives: to estimate survival, after AIDS diagnosis, in people who got infected with HIV through injecting drug use (IDUs), to identify among variables collected at AIDS diagnosis those which were associated to prognosis and to assess the frequency of morbid conditions at death. Design: population-based, longitudinal study. Setting and participants: 4,040 IDUs diagnosed with AIDS in Italy between 1999 and 2005. Methods: vital status tip to 2006 was retrieved through a record-linkage procedure with italian mortality database. Kaplan-Meier method and multivariate Cox model were used to estimate survival curves and compute hazard ratios of death (HR), and corresponding 95% confidence intervals (95% CI), for several prognostic factors, respectively. Results: the 2-year and 5-year survival probabilities after AIDS diagnosis of IDUs were 72% and 60%, respectively. Elevated risks of death emerged for IDUs with older ages (HR=2.0 95% CI 1.6-2.4 for >= 45 years old vs. < 35 years old), lower education (HR=1.4 95% CI 1.2-1.7 for elementary school vs. high school/university), longer time span between first HIV positive test and AIDS diagnosis (HR=1.6 95% CI 1.4-1.9 for >= 6 months vs. < 6 months), and lower CD4 cell count at diagnosis (HR=1.5 95% CI 1.3-1.7 for <50 cells/mm(3) vs. >= 200 cells/mm(3)). Compared to Pneumocystis carinii pneumonia, non-Hodgkin lymphomas were the worst prognostic factors, particularly primary brain lymphoma (HR=7.2, 95% CI 4.4-11.8). Over 1,581 deceased IDUs, for 1,567 cases death certificates were available. 52% of cases reported no AIDS-defining illnesses: 64 (4%) violent causes, 94 (6%) cancers, and 656 (42%) only non neoplastic illnesses, among which 415 (27%) liver diseases. Conclusion: the results of this population-based study showed that, in the highly active antiretroviral therapy era, survival of IDUs with AIDS was still lower compared to that of HIV sexual transmission groups. The presence at death, in 52% of cases, of non AIDS-defining illnesses indicates the important role oil mortality of co-morbidities, including liver diseases and violent causes. (Epidemiol Prev 2009; 33 (4): 184-89)

Objectives: to estimate survival, after AIDS diagnosis, in people who got infected with HIV through injecting drug use (ID Us), to identify among variables collected at AIDS diagnosis those which were associated to prognosis and to assess the frequency of morbid conditions at death. Design: population-based, longitudinal study. Setting and participants: 4,040 ID Us diagnosed with AIDS in Italy between 1999 and 2005. Methods: vital status up to 2006, was retrieved through a recordlinkage procedure with Italian mortality database. Kaplan-Meier method and multivariate Cox model were used to estimate survival curves and compute hazard ratios of death (HR), and corresponding 95% confidence intervals (95% CI), for several prognostic factors, respectively. Results: the 2-year and 5-year survival probabilities after AIDS diagnosis of ID Us were 72% and 60%, respectively. Elevated risks of death emerged for IDUs with older ages (HR=2.0 95% CI 1.6-2.4 for ≥45years old vs. <35years old), lower education school/university), longer time span between first HIV positive test and AIDS diagnosis (HR=1.695% CI 1.4-1.9 for ≥ 6 months vs. <6months), and lower CD4 cell count at diagnosis (HR= 1.5 95% CI 1.3-1.7for <50 cells/mm3 vs.≥ 200 cells/mm3). Compared to Pneumocystis carinii pneumonia, non-Hodgkin lymphomas were the worst prognostic factors, particularly primary brain lymphoma (HR=7.2,95% CI 4.4-11.8). Over 1,581 deceased IDUs, for 1,567 cases death certificates were available. 52% of cases reported no AIDS-defining illnesses: 64 (4%) violent causes, 94 (6%) cancers, and 656 (42%) only non neoplastic illnesses, among which 415 (27%) liver diseases. Conclusion: the results of this population-based study showed that, in the highly active antiretroviral therapy era, survival of IDUs with AIDS was still lower compared to that of HIV sexual transmission groups. The presence at death, in 52% of cases, of non AIDS-defining illnesses indicates the important role on mortality of co-morbidities, including liver diseases and violent causes.