Increased regression and decreased incidence of human papillomavirus- related cervical lesions among HIV-infected women on HAART

Objective: To determine the impact of HAART on incidence, regression, and progression of cytopathological abnormalities in HIV-infected women. Design: Prospective cohort. Methods: HIV-infected women (N=1123) from Soweto, South Africa underwent serial cervical smears that were analyzed and reported using the Bethesda System. The results of HAART and non-HAART users were compared using two statistical approaches: a survival analysis assessing risk of incident smear abnormality among women with baseline normal smear results; and analysis with marginal models assessing for an association between HAART use and likelihood of regression/progression in consecutive smears. Results: After multivariate survival analysis, women using HAART with a normal baseline smear were 38% less likely to have an incident smear abnormality during follow-up than nonusers [confidence interval (CI) 0.42-0.91; P=0.01]. Multivariate marginal models analysis identified a significantly increased likelihood (odds ratio 2.61; CI 1.75-3.89; P