MORTALITY AND SURVIVAL TRENDS IN PATIENTS WITH AIDS IN NORTH-EAST ENGLAND FROM 1984-1992

Objective: to study trends in mortality and survival in patients with AIDS attending an ID unit. Method: retrospective analysis of patients developing an AIDS-defining illness between April 1984, and November 1 992. Survival was analysed by calculation of survival product-limit. R esults: 71 patients were analysed (including four women), 23 of whom a re still alive. Pneumocystis carinii pneumonia (PCP) was the most freq uent AIDS-innex diagnosis: n = 36 (51%); 24 of these patients have die d. HIV encephalopathy was the most frequent diagnosis at death; n = 16 (22.5%), followed by mycobacterial infection; n = 11 (15.5%), and PCP and CMV infection, each occurring in 10 (24%). One-, 2- and 3-year su rvival probabilities for patients with AIDS before 1987 were 0.46, 0.1 5 and 0 compared with probabilities of 0.63, 0.5 and 0.3 in those diag nosed after 1987; log rank -P <0.01. One- and 2-year survival probabil ities in patients who received at least 3 months' zidovudine (AZT) the rapy were 0.76 and 0.53 in those who are still alive compared with 0.5 5 and 0.33 in the deceased, while values for deceased AZT-naive patien ts were 0.29 and O.1; -P <0.O1, Thirteen (27%) deaths occured within 2 months of an AIDS-index disease. In 10 patients this was their first presentation to the department. PCP accounted for 8 (61%) of these dea ths. Conclusions: survival in patients with AIDS has increased since 1 987, when AZT was introduced. Early AIDS-related deaths are frequent i n patients who have had no prior medical care. This has implications f or education and provision of care in individuals with asymptomatic HI V infection.