My 25th year at UMBC marked my second transition during my time at the university … first from a student to a member of the staff, and now to a faculty member in the Information Systems department. In that time I have completed four academic degrees, collaborated with many brilliant colleagues, and accomplished more in my professional career than I thought possible. I was formerly the Senior Director of Analytics and Research at The Hilltop Institute, where I focus on the quantitative analysis of a diverse array of healthcare data sources. By applying sophisticated analytical methods to these data, I helped clients build more effective and efficient health care programs while sharing what I learn with the research community. I was also Hilltop’s Chief Data Scientist, overseeing our methodological approaches to public health interventions, social drivers of health, and healthcare financing. My focus area is creating predictive models to assess the probability of adverse events among at-risk populations, including older adults and individuals with disabilities. I have served as principal investigator or project director for projects funded by many partners, including the Robert Wood Johnson Foundation, the Commonwealth Fund, the National Science Foundation, the National Institutes for Health, and multiple state agencies. I also have an affiliate faculty appointment with the Erickson School of Aging Studies, where I help students use data science to care for an aging population.
I have dedicated my career to helping people who need help the most. We have statistical models available that can quantify such need, allowing us to prioritize scarce health care resources to individuals with the highest risk. With the correct methods and diligent design, these models can predict adverse health events with high accuracy. I have led the development of two such models currently in production health information exchange environments: one prioritizing the waiting list of potential Medicaid home- and community-based services participants, and the other guiding the care coordination and outreach activities for hundreds of primary care practices participating in the Maryland Primary Care Program. Together these models serve over 400,000 individuals. They were guided by extensive stakeholder input from patients, caregivers, providers, and administrators. These groups each gave a unique perspective on how the models could best drive care and were invaluable in the development process. Other models currently under development predict hospitalization due to COVID-19, hospital readmissions, and repeat violent injury. I have also received support from the National Science Foundation to examine the fairness and equity of resource distributions driven by our algorithms. I served as primary investigator, project director, or co-PI on all these projects
Aligning with my focus on vulnerable populations, I have done extensive work monitoring multiple state Medicaid programs. Because Medicaid covers 1 in 5 Americans and has significant state-by-state variability, it is important that each Medicaid implementation is monitored to ensure that it serves recipients as effectively as possible. I have developed and produced multiple program development and evaluation analyses for Maryland, New Mexico, Rhode Island, and New Jersey using a mix of Medicaid administrative data, nursing home assessments, and stakeholder interviews. I have also participated in research identifying the incidence of violent injury in Maryland’s Medicaid population. Particular attention has been paid to projects specific to the aged and disabled population, which led to an appointment with the CMS Medicare-Medicaid Coordination Office examining the characteristics of dually eligible individuals. I have also led the development of a metrics-based analysis of Maryland’s Money Follows the Person program and other long-term care rebalancing initiatives and lead an analysis to estimate the fiscal effect of Medicaid expansion in Mississippi.
While interventions aimed at helping vulnerable populations are commendable, it is important to rigorously scrutinize the effects of those programs to ensure that they benefit the target population. I have performed many such analyses, focusing both on specific intervention groups as well as larger cohorts of at-risk individuals. I built the propensity-matched comparison group, extracted Medicaid data files, and built analytical Medicare data files for use by Johns Hopkins in the evaluation of the Maximizing Independence at Home (MIND) and Community Aging in Place-Advancing Better Living for Elders (CAPABLE) programs, both implemented in an older adult population in Maryland. I also completed an analysis of Maryland’s Community First Choice program, highlighting the changing mix of participants and their corresponding care needs. During each evaluation I was privileged to engage with various stakeholder groups, which allowed me to understand the broader context surrounding each intervention and gave me multiple audiences for my findings.