Research

My research focuses on the design and analysis of clinical trials with correlated observations, as well as the use of interim monitoring in randomized trials and observational studies. A guiding tenant of my research philosophy is that new biostatistical methods should be useful in realistic biomedical research scenarios; thus I work closely with my clinical collaborators, particularly in the application area of Alzheimer's disease. Below are brief descriptions of some of my recent projects and collaborations.

Pragmatic and Cluster Randomized Trials

Cluster randomized trials are studies where treatment is randomized at the cluster level but outcomes are typically collected at the individual level. This branch of trials can also include pragmatic modifications such as stepped-wedge trials, which progressively transition clusters from the control to the intervention condition and the transition timing is randomized for each cluster. Specific projects include the optimal design of cluster randomized trials with heterogeneous treatment effects in the presence of uncertainty around ICCs, development of power calculations for stepped-wedge trials with time-to-event endpoints, and systematic reviews of current stepped-wedge trials in the health field.

Sequential Monitoring

Interim monitoring or group sequential procedures are commonly used to efficiently conduct randomized trials of therapeutic interventions, but there is utility beyond these settings. . Specific projects include the integration of interim testing into two-stage patient preference designs, the use of adaptive sequential monitoring in observational studies, and investigations into the use of kappa statistics in sequentially monitored trials.

Alzheimer's disease

Human studies of Alzheimer's disease are unique in that they require the recruitment of two individuals: the primary study participant and the study partner that assists with study compliance and completes outcomes assessing participant cognitive and functional ability. To adequately understand how to design AD studies, and recruit and retain participants, we must understand the motivations of both study participants as well as their study partners. Specific projects include exploration of factors influencing preclinical trial enrollment, the role of study partners in studies of Alzheimer's dementia and preclinical AD, and the effect of study partner relationship and turnover on endpoint estimation in AD trials.