CD4 Cell Count Distributions and HIV‐Associated Mortality in Africa

Abstract

Lawn and Wood raise an important issue concerning the relationship between CD4 cell counts and mortality in people infected with HIV and note that there is substantial variation in CD4 cell counts at death. As CD4 cell counts decline the incidence of opportunistic infections increases monotonically; if such infections are left untreated then people may die when their CD4 cell counts are still relatively high just as an HIV-negative person might. For example in a comparison of 8 studies the mean CD4 cell count in HIV-positive patients presenting with tuberculosis was 202 cells/µL (range 136-269 cells/µL). If better data on the time course of CD4 cell counts in a sufficiently large sample of HIV-positive patients were available more-sophisticated models of the relationship between CD4 cell count decline the incidence of opportunistic infections and mortality could be developed. Our model which draws on the limited data that are currently available is a step in this direction. It is nevertheless worth noting that evidence from industrialized as well as low- and middle-income countries suggests that the majority of AIDS-related deaths occur at very low CD4 cell counts. For example in a study in South Africa in the early 1990s the median CD4 cell count at the onset of AIDS was 98 cells/µL for heterosexual patients and 40 cells/µL for homosexual patients after which the median survival time was 17 and 7 months respectively. In an Australian cohort followed between 1986 and 1991 the median CD4 cell count at death was 10 cells/µL. In Uganda a recent study showed that the median CD4 cell count at death was 24 cells/µL. In the United Kingdom a study reported mean CD4 cell counts at death of 19 cells/µL in 1988 44 cells/µL in 1997 and 58 cells/µL in 1998. (excerpt)