SURVIVAL OF AIDS PATIENTS ACCORDING TO TYPE OF AIDS-DEFINING EVENT

Citation
Aj. Mocroft et al., SURVIVAL OF AIDS PATIENTS ACCORDING TO TYPE OF AIDS-DEFINING EVENT, International journal of epidemiology, 26(2), 1997, pp. 400-407
Citations number
28
Categorie Soggetti
Public, Environmental & Occupation Heath
ISSN journal
03005771
Volume
26
Issue
2
Year of publication
1997
Pages
400 - 407
Database
ISI
SICI code
0300-5771(1997)26:2<400:SOAPAT>2.0.ZU;2-O
Abstract
Background. There are known to be wide differences in the prognosis of patients with a diagnosis of AIDS, In this study of 6578 patients wit h AIDS from 17 European centres, we develop a ranking of AIDS-defining illnesses, and determine how well this ranking holds alter adjustment for potential confounding variables, Methods, Survival from each AIDS -defining event was calculated and ranked using Kaplan-Meier estimatio n of median survival, Cox proportional hazards models with each diseas e modelled as a time dependant covariate were used to determine the ri sk of death after each diagnosis, before and after adjustment for pote ntial confounders, Results. Median survival after an initial AIDS-defi ning diagnosis of progressive multifocal leukoencephalopathy and malig nant lymphoma was particularly poor (2 and 5 months respectively), whi le tile longest median survival occurred alter initial AIDS-defining i llnesses of Kaposi's sarcoma and extrapulmonary tuberculosis (17 and 2 2 months respectively). Patients diagnosed with a primary brain lympho ma had shorter median survival times than patients with a peripheral l ymphoma (median survival of 1 month and 4 months respectively, P<0.000 1). In general, median survival in patients with cutaneous Kaposi's sa rcoma (skin, oral) was between two and four times longer than patients with systemic involvement. The ranking of diseases was found to be ge nerally similar after adjustment for all potential confounders. Conclu sions. AIDS-defining events can be grouped into three categories with median survival alter diagnosis of <6 months, 6-12 months and >12 mont hs, The assigned rankings oi disease would not he altered by prognosti c factors such as age or CD4 lymphocyte count. These results have impo rtant implications in the design or clinical trials and patient manage ment.