The aim of this study was to investigate the mortality rate, risk factors and causes of death among HIV-infected patients in Guadeloupe from 1988 to 2009. We used Kaplan-Meier analysis to describe the survival trends and the Cox proportional hazard model to identify predictors of deaths in HIV-infected patients. Mortality rate and causes of death were compared among patients whose HIV diagnosis was made in two different study periods. There were 672 deaths recorded. The exact cause of death was clearly identified for 202 patients (35%). There were 672 deaths recorded. The exact cause of death was clearly identified for 202 patients (35%). There were 165 AIDS defining events and 37 non-AIDS defining events. The most frequent causes of death reported were HIV encephalopathy (n=54), cerebral toxoplasmosis (n=39), HIV-wasting syndrome (n=37), malignancies (n=13), cytomegalovirus pneumopathy (n=10). The crude incidence rate for patients on HAART was between 2.2 and 3.1 per 100 person-years, whereas it was 17.2 for those not on HAART. The variables in the Cox model that were significantly associated with death were addiction, depression, low CD4 count and high viral load levels in baseline. These results suggested that increased screening efforts, particularly in marginalised populations could help with early diagnosis and better follow-up in order to reduce mortality among HIV-infected patients.
Manuscripts that are Published Ahead of Print have been peer reviewed and accepted for publication by the Editorial Board of the West Indian Medical Journal. They may appear in their original format and may not be copy edited or formatted in the style guide of this Journal. While accepted manuscripts are not yet assigned a volume, issue or page numbers, they can be cited using the DOI and date of e-publication. See our Instructions for Authors on how to properly cite manuscripts at this stage. The contents of the manuscript may change before it is published in its final form. Manuscripts in this section will be removed once they have been issued to a volume and issue, but will still retain the DOI and date of e-publication.