Survival in HIV infection: do sex and category of transmission matter? Swiss HIV Cohort Study.

OBJECTIVE

To examine survival differences in HIV-infected individuals between the sexes and the categories of transmission.

PATIENTS AND SETTING

Subjects assumed to have been infected by injecting drug use, by male homosexual contact or by heterosexual contact and enrolled in the Swiss HIV Cohort Study, a national multicentre cohort study of adult HIV-infected patients.

METHODS

Kaplan-Meier lifetable and Cox regression analyses were performed (time 0 = date of study entry) and survivors were censored at their last follow-up visit.

RESULTS

A total of 4428 patients (mean duration of follow-up, 2.1 years) were considered. At entry, men were older and had lower CD4+ counts than women (P < 0.0001). Homosexual men were the oldest group with the lowest CD4+ counts and the most advanced disease at study entry (P < 0.0001). Crude hazard ratios indicated a 28% lower mortality from all causes in women compared with men (P < 0.0001) and, among men but not in women, a 28 and 32% higher mortality in homosexuals and heterosexuals compared with injecting drug users (IDU) (P = 0.0002 and P = 0.008, respectively). After adjusting for differences at entry, the mortality difference between the sexes disappeared (P = 0.5) and differences across transmission categories were reversed. Mortality in homosexual men was an estimated 13% (P = 0.057) lower than in male IDU; mortality was 22% lower in heterosexual women than in female IDU (P = 0.098). In stratified analysis the increased risk in IDU was confined to subjects with CD4+ cell counts > 500 x 10(6)/l cells at study entry.

CONCLUSIONS

These results indicate a uniform mortality risk across the sexes but indicate an increased risk in IDU without CD4+ cell depletion at entry, the latter probably attributable to causes not related to progression of HIV infection. This study underscores the importance of adjusting for prognostic factors when comparing survival between different patient groups.