13th Conference on Retroviruses and Opportunistic Infections Denver, Colorado - February 5-8, 2006

Conf Retrovir Opportunistic Infect 2006 Feb 5-8;13:abstract no. 30

Philip J Peters^1,2, I Zulu^1,3, N Kancheya^1,3, S Lakhi^1,3, E Chomba^1,3, D J Kim^1,4, I Brill^1,4, J Meinzen-Derr^1,4, A Fraser-Bell^1,5, and S Allen^1,5,6
1Rwanda-Zambia HIV Res Group; ²Emory Univ Sch of Med, Atlanta, GA, US; ³Univ Teaching Hosp, Lusaka, Zambia; ⁴Univ of Alabama at Birmingham, US; ⁵Emory Rollins Sch of Publ Hlth, Atlanta, GA, US; and ⁶Emory AIDS Intl Training and Res Prgm, Atlanta, GA, USA

BACKGROUND: Mortality surveillance in HIV-infected people remains a vital public health tool, however, for more than 90% of Africa’s population there is no information on causes of adult mortality. We evaluated rates of survival and causes of mortality in the largest prospective, community-based cohort of HIV discordant couples in Africa.

METHODS: 1528 HIV discordant couples (3056 people) were recruited from couples’ VCT centers in Lusaka, Zambia between January 1995 and December 2003 as part of a study on heterosexual HIV transmission and followed at 3 month intervals. Subjects who initiated anti-retroviral therapy (ART) were censored from analysis. Mortality rates were compared by HIV status, sex, CD4 count, viral load, 2005 WHO stage, and Modified Kigali Combined stage (MKC stage– WHO-based staging system modified to incorporate ESR, HCT, and BMI). Kaplan-Meier survival and Cox proportional hazard methods were used to calculate time to mortality and relative hazards. Cause of death was ascertained by verbal autopsy, chart review, and death certificate.

RESULTS: From 1995 to 2004, 392 people died (12.8% of cohort) over 10,378 person years (py) of follow-up. HIV positive individuals had a median estimated survival of 8.9 years (95% CI=8.2 - 9.1 yrs) from enrollment. Mortality rates for HIV-infected men and women were 8.8 and 6.0 per 100 py. The 3-year mortality rates for individuals with MKC stage 1, 2, 3, and 4 disease at enrollment were 12.7%, 11.4%, 24.7%, and 51.3% respectively. MKC stage 4 disease (univariate HR 5.5, 95% CI=3.7-8.2) was a stronger predictor of mortality than 2005 WHO stage 4 disease and was comparable to CD4 count <200/mm³ and viral load >5log copies/mL. Tuberculosis and chronic gastroenteritis were the primary causes of death among HIV positive Zambians accounting for 24% and 20% of mortality respectively. Traditional AIDS-defining illnesses were less frequent but still significant causes of mortality with Kaposi's sarcoma, cryptococcal meningitis, and candidal esophagitis accounting for 6.3% of deaths combined.

CONCLUSIONS: HIV positive Zambians had comparable survival times to pre-ART cohorts in high-income countries. MKC staging is a powerful, low-cost tool to identify people at high risk for death who need urgent evaluation for ART. The burden of mortality due to tuberculosis and chronic gastroenteritis highlights the need for parallel investment in tuberculosis and diarrheal disease treatment as ART scales up.

Download abstract

2006-02-05
30

Copyright © 2006 - Foundation for Retrovirology and Human Health. Reproduction of this abstract (other than one copy for personal reference) must be cleared through the Foundation for Retrovirology and Human Health.