Developing novel statistical methods to analyze longitudinal and time-to-event data in survivorship studies and analyzing childhood cancer survivor data
The data acquired in observational, long-term follow up studies like SJLIFE and CCSS are self-reported, and often, imperfect. Nevertheless, investigators rely on this data to make decisions about clinical care. Our team is working to develop statistical methods to improve the accuracy of this data and implement those methods in collaborative projects to guide analysis and interpretation. We have a particular interest in neurological and cardiological chronic health conditions, among others, that arise in childhood cancer survivors following treatment. Our work is conducted in collaboration with the Department of Epidemiology and Cancer Control contributes robustly to the mission of St. Jude.
Our lab is focused on the analysis and interpretation of childhood cancer survivor data within the St. Jude LIFE Study and the Childhood Cancer Survivor Study (CCSS). The goal of these studies is to determine whether there are associations between early therapeutic treatments and long-term sequelae experienced by survivors. The data collected are observational, and reliant on patient self-reporting of clinical events and timelines. We develop statistical methodologies to reduce the inherent bias that exists in these recall-based data and analyze them in a way that we can ensure the most accurate interpretation possible.
Our research interests cut across different areas of statistics, including survival and event history analysis, modeling, theory and methods for estimation, prediction, and the analysis of incomplete data. We are motivated by scientific and technical issues that arise in medicine, public health, and currently cancer survivors. Part of our scientific goal focuses on joint models for longitudinal and time-to-event data with applications in biomarker identification and precision medicine.