AEGiS-13IAC: Comparison of medhods to determine the effectiveness af antiretroviral therapies (ART) at the population level in HIV-1 incident and prevalent cohorts.

13th International AIDS Conference

Durban, South Africa - July 9-July 14, 2000

Comparison of medhods to determine the effectiveness af antiretroviral therapies (ART) at the population level in HIV-1 incident and prevalent cohorts.

Int Conf AIDS 2000 Jul 9-14; 13:(abstract no. ThOrC746)

Tarwater PM, Munoz A, Mellors J, Gore ME, Margolick JB, Phair J, Detels R
P.M. Tarwater, Dept. Epidemiology JHU Sch of Public Health, 615 N Wolfe Street Rm E7139, Baltimore, MD 21205, United States, Tel.: +1 410 955 4320, Fax: +1 410 955 7587, E-mail: ptarwate@jhsph.edu

OBJECTIVE: In analyses of the effectiveness of ART in the Multicenter AIDS Cohort Study (MACS), we compared the results obtained by two different

METHODS: adjusting for the duration of HIV infection vs. adjusting for baseline disease markers (HIV-1 RNA and CD4 cell count).

METHODS: After characterizing treatment eras by time periods (Jan 1986-Jun 1989 = no ART, Jul 1989-Dec 1993 = mono NRTI ART, Jan 1993-Jun 1996 = combination NRTI ART, and Jul 1996-Jun 1999 = potent ART), the two methods of adjustment for disease progression were analyzed by comparing relative hazards estimated from Cox proportional hazard models. The first method requires a seroconverter cohort whereby using duration of HIV infection as the time scale, relative hazards are estimated at equal infection duration. The second method uses measures of CD4 cell count and HIV-1 RNA level, measured at the beginning of each era, as covariates and uses time in each era as the time scale, thus allowing for use of prevalent cohorts.

RESULTS: The relative hazards of AIDS after an equal duration of HIV infection were 1.52, 0.91, and 0.30 for the eras of no ART, combination NRTI ART, and potent ART respectively (using the era of mono NRTI ART as the reference). The alternative analysis adjusting for both CD4 cell count and HIV-1 RNA level, measured at the beginning of each time period, showed very similar relative hazards (1.52, 1.03, and 0.31, respectively). The inferences were identical in that potent ART was highly significant (p > 0.001), no ART was borderline significant, and combination NRTI ART was not statistically different from the mono NRTI ART.

CONCLUSION: We have shown that in analyses of the effectiveness of ART in both incident and prevalent cohorts two different methods of adjustment for risk of disease progression provide similar results. Since most HIV studies do not have information on the date of seroconversion, adjustments for CD4 cell count and HIV-1 RNA level as markers for disease progression should prove useful in analyzing population-level effectiveness of ART.

Keywords: AEGIS, HIV-1, CD4 Lymphocyte Count, HIV Infections, HIV Seropositivity, Acquired Immunodeficiency Syndrome, Disease Progression, Antiretroviral Therapy, Highly Active, Anti-HIV Agents, HIV, HIV Antibodies, Proportional Hazards Models, HIV Seronegativity, Drug Therapy, Combination, immunology, therapy, drug therapy
000709
ThOrC746