My research is focused on issues related to the design and analysis of studies of multi-component interventions and studies of multifactorial geriatric health conditions and multiple chronic conditions. I develop strategies for handling missing data or competing risks which result in the primary outcome not being observed that frequently occurs in studies of older persons. I am the founder of “Gerontologic Biostatistics” that focuses on statistical and epidemiologic methodology development in aging research. This discipline trains biostatisticians for conducting collaborative research with clinical investigators in geriatrics and gerontology and provides the basis for the development of new statistical methodologies. Methods include: longitudinal modeling of trajectories, extended Cox models for state transitions in geriatrics, such as frailty and disability that consist of multiple discrete states in which both onset and recovery are possible, and determining the mechanisms of action of an effective multi-component intervention. Most recently, the use of joint modeling to account for deaths in longitudinal studies of geriatric outcomes. We have developed a website GRASP that aids all who are interested in Aging Research, both those familiar with analytic methods and those getting familiar with many of our methods.

Design and analysis of studies of multi-component interventions and studies of multifactorial geriatric health conditions and multiple chronic conditions.