Deaths in the era of HAART: contribution of late presentation, treatment exposure, resistance and abnormal laboratory markers.

OBJECTIVES: To describe the characteristics of deaths that occur among HIV-positive individuals in the HAART era. DESIGN: Observational database. METHODS: Deaths were reviewed that occurred among HIV-positive individuals seen at the Royal Free Hospital, London between January 1998 and December 2003. RESULTS: Over the study period, there were 121 deaths; death rates declined from approximately 2.0/100 person-years of follow-up in 1998-2000 to approximately 1.0/100 person-years of follow-up in 2001-2003. Approximately one half of deaths (45.5%) were from AIDS-related causes and 74 deaths (61.2%) occurred in individuals who had received HAART: patients had been exposed to a median of seven (range 2-14) antiretroviral drugs and two-fifths had started treatment in the pre-HAART era. Another 15 patients had received only non-HAART treatment regimens prior to death. The median pre-death CD4 cell counts were 68 and 167 cells/microl among those who had and had not received HAART; 23 (31.1%) and 4 (8.5%) had HIV RNA < 400 copies/ml, respectively. Of the patients exposed to HAART for at least 6 months and who experienced viral rebound, information was available on resistance for 26 (21.5% of the total deaths) and 19 of those tested had at least one resistance mutation (median 5, range, 1-16). CONCLUSIONS: While mortality rates among HIV-infected individuals at our centre have fallen since 1988, the deaths that do now occur are more diverse and are the result of a number of factors, including late presentation, delayed uptake of HAART and the previous use of treatment combinations that are now viewed as suboptimal.