Objective: To describe the epidemiological, clinical and economic changes that occurred in the HIV epidemic in Italy prior to and after the introduction of highly active antiretroviral therapy (HAART).

Design: A prospective, observational, multicentre case-control study was conducted comparing data, collected over 6 months, from an AIDS cohort in 1998 with that of a cohort in 1994. Out of 77 patients with AIDS in the 1998 cohort, 74 survived. These 74 patients were matched for severity of illness with 74 patient survivors from the 1994 cohort to enable valid comparisons of mortality, disability-dependency (DD), health-related QOL (HR-QOL), and direct costs.

Results: Overall, a considerable difference was observed in mortality (33.8% in 1994 vs 3.9% in 1998) between unmatched patients of the two cohorts. As for matched patients, the number of hospital admissions was 1.7 in 1994 and 0.8 in 1998; the average length of stay was 28.1 days in 1994 and 12.6 days in 1998. The direct cost per patient per year was €15 390 and €11 465 for the 1994 and 1998 cohorts, respectively (1999 values). The 1998 patient cohort had significantly better HR-QOL at 6 months in two domains of the instrument used (emotional reaction and energy) and the percentage of totally dependent patients was significantly lower compared with the 1994 cohort (1.4% vs 6.8%).

Conclusions: This is the first study to present a comprehensive comparison of direct costs, DD and HR-QOL of patients with AIDS between two time periods. The use of a case-control design has enabled changes in costs and outcomes to be linked to the introduction of HAART in Italy in 1997.