Survival of patients with the acquired immunodeficiency syndrome in Australia

Abstract

The duration of survival of all persons with the acquired immunodeficiency syndrome (AIDS) in Australia who were diagnosed to July 31, 1987 was analysed by means of standard Kaplan-Meier methods in order to construct survival curves, and by log-rank tests which were used to assess the significance of difference among the curves. The median duration of survival was 10.4 months. Significant (P<0.001) differences in the survival curves were found between men and women (median survival, 11.4 and 3.8 months, respectively) by the log-rank test. Patients who were aged 20-49 years at diagnosis had a longer (P<0.001) median survival (11.4 months) than did those who were older than 60 years of age (2.7 months) and those who were aged less than 20 years (0.5 months) at diagnosis. Those patients who were infected by way of blood transfusions had a shorter (P<0.001) median survival (1.6 months) than did those who were infected through homosexual activity (11.4 months). Median survival time, which was stratified by the disease stage at presentation, varied from 12.4 months in those patients with Kaposi''s sarcoma alone to 7.0 months in those patients with lymphoma (P=0.01). The median survival time for those patients who presented solely with Kaposi''s sarcoma was shorter than was that which has been reported from other developed countries. Multivariate analysis of all male cases by means of the proportional-hazards model indicated that the presentation of cases at an age that was greater than 50 years (relative risk, 2.36; 95% confidence interval [CI], 1.63-3.42) was associated independently with a reduced survival time, while the presentation of cases solely with Kaposi''s sarcoma (relative risk, 0.67; 95% CI, 0.49-0.92) was associated independently with an increased survival time. When all cases were included in this analysis, these same predictors of survival were significant; in addition, cases of AIDS as a result of the receipt of a blood transfusion were associated independently with a reduced survival time (relative risk, 1.62; 95% CI).