Objectives: To investigate temporal changes in survival of persons with AIDS (PWA) diagnosed in Lazio, Italy.
Design: Longitudinal study of 2862 PWA registered at the AIDS Surveillance System in the period 1 January 1993 to 30 June 1997 and followed for vital status up to 31 May 1998.
Methods: Hazard ratios (HR) of death were calculated by year of diagnosis and by year of follow-up, adjusting for gender, age, modality of exposure, CD4 cell count at diagnosis, and AIDS-defining illness, using the Cox proportional regression model.
Results: In the period 1 January 1993 to 31 May 1998, 1914 PWA died. The multivariate model showed a decrease of the risk of death for patients diagnosed from 1995 onward compared with patients diagnosed in 1993 [1995: HR 0.82; 95% confidence intervals (CI) 0.73-0.93; 1996: HR 0.44; 95% CI 0.38-0.51; first semester of 1997: HR 0.47; 95% CI 0.37-0.59]. The model with the year of follow-up modeled as time-dependent variable showed a statistically significant decline in the risk of death starting from the first quarter of 1997 (HR 0.62; 95% CI 0.48-0.80) and reached the minimum in the first two quarters of 1998 (first quarter: HR 0.13; 95% CI 0.08-0.21; second quarter: HR 0.14, 95% CI 0.08-0.25). There was a small difference in the magnitude of the risk estimates among genders, with a slower improvement in survival among females [first semester of 1997 for males HR was 0.51 (95% CI 0.39-0.66) whereas for females HR was 0.80 (95% CI 0.51-1.27)].
Conclusions: Survival of PWA in a population-based study improved significantly in the era when highly active antiretroviral therapy became available, but such improvement is not homogeneous among genders.