As an associate professor of medicine in the Division of Population Health, my academic pursuits are centered on the intersection of data science, artificial intelligence, and wearable technologies to improve health outcomes for individuals living with Type 2 diabetes mellitus. My work is dedicated to developing innovative, data-driven solutions that leverage wearable devices to predict and manage glycemic control, ultimately aiming to transform diabetes care through real-time monitoring, predictive analytics, and tailored interventions. By bridging the latest advances in technology with practical healthcare applications, my research seeks to make meaningful, evidence-based improvements in diabetes management, thereby addressing a pressing public health challenge.

Building on this focus, my prior background in clinical trials, where I conducted significant work in bone marrow transplant studies, provides an essential foundation for my current work in diabetes care. Notably, as part of a collaborative team, I contributed to a phase 3, multicenter, randomized controlled trial that examined the efficacy of cyclophosphamide-tacrolimus-mycophenolate mofetil, an experimental prophylaxis regimen, in comparison with the standard regimen of tacrolimus-methotrexate for preventing graft-versus-host disease (GVHD) in patients undergoing their first allogeneic hematopoietic stem cell transplantation for hematologic cancers. Among 431 adults receiving peripheral-blood grafts from HLA-matched related or matched/mismatched unrelated donors, the experimental regimen yielded a significantly higher rate of GVHD-free, relapse-free survival at one year. As a result, cyclophosphamide-tacrolimus-mycophenolate mofetil has become the new standard of care for GVHD prophylaxis in adults undergoing reduced-intensity conditioning allogeneic transplants, marking a critical advancement in this field.

In addition to my clinical trials work, complex sample surveys play a crucial role in healthcare by providing comprehensive data that informs policy-making, resource allocation, and health interventions at both population and individual levels. These surveys often involve intricate design features such as stratification, clustering, and unequal probability sampling, posing challenges for accurate data analysis and inference. I have made substantial contributions to the field of complex sample surveys, with a particular emphasis on improving the precision of variance estimation. Traditional ordinary least squares mean regression methods often struggle with highly skewed outcomes, leading to biased and inefficient parameter estimates. To address this challenge, I have developed and applied median regression techniques specifically designed for complex sample surveys, providing a more robust and reliable method for analyzing skewed outcomes—a common characteristic in survey datasets. This work ensures accurate statistical inference and strengthens the utility of complex survey data for generating actionable, policy-relevant insights across multiple sectors.

Together, these threads of research reflect my commitment to leveraging data-driven approaches to enhance health outcomes, bridge technological innovation with healthcare practice, and improve the accuracy and utility of data-driven insights for better public health decision-making.