Department of Biostatistics Seminar/Workshop Series

Estimation of the standardized risk difference and ratio in a competing risks framework: Application to injection drug use and progression to AIDS after initiation of antiretroviral therapy.

Stephen Richards Cole, PhD

Professor of Epidemiology, University of North Carolina, Gillings School of Global Public Health

Wednesday, September 25th, 1:30-2:30pm, MRBIII Room 1220

---

We present an example estimating the effect of injection drug use on the 6-year risk of AIDS incidence after initiation of combination antiretroviral therapy in an 8-site US clinical cohort study with all-cause mortality as a competing risk. We estimate the risk standardized to the total study sample using inverse probability weights and account for the competing risk of mortality using the cumulative incidence function; estimates of precision are obtained by nonparametric bootstrap. In the sample of 7182 patients (83% male, 33% African American, median age 38 years), we observed a 6-year standardized AIDS risk of 16.75% among 1143 injection drug users and 12.08% among 6039 non-users, yielding a standardized risk difference of 4.68 (95% CI: 1.27, 8.08) and a standardized risk ratio of 1.39 (95% CI: 1.12, 1.72). Results may be sensitive to the assumptions of exposure-variation irrelevance, no exposure measurement bias, and no unmeasured confounding. These limitations suggest that results be replicated with more refined measurement of injection drug use.

-- ColeBeck - 16 Jul 2013