Professor of Biostatistics
Other Affiliations
Research
Dr. Lagakos' current research involves a variety of statistical issues arising in clinical trials and other longitudinal studies, with particular emphasis on statistical methods and analyses relating to HIV and other infectious diseases.
One area of current interest in statistical methods for estimating the timing of mother-to-child transmission of HIV, which can occur in utero or after birth through breast feeding. The statistical challenge arises because in utero infection status cannot be directly observed, the diagnostic tests given to newborns are imperfect and periodic, and the duration of breast-feeding (and hence cumulative risk of infant infection) may depend on the intervention.
Another area of current interest is the estimation of HIV incidence rates using cross-sectional samples using a battery of diagnostic tests of differing sensitivity. In many settings, the traditional approach of conducting a longitudinal study to estimate HIV incidence is not practical. The cross-sectional sample approach aims to identify persons with a 'recent infection' (defined as testing positive on the more-sensitive diagnostic test and negative on the less-sensitive test) and then to use this along with external information to estimate the incidence rate. Although potentially very useful, the approach suffers from methodological problems that must be remedied before it can be reliably used to assess levels and trends in HIV incidence.
0 comments:
Post a Comment