Health Services Research

Our health services research team uses rigorous and innovative analytic methods to study the delivery of health services, using large administrative databases including Medicare, SEER-Medicare, Optum, the Texas Department of Criminal Justice database, and the University of Texas electronic health record (EHR) data.

Our investigative team has focused on range of important topics including opioid prescribing and adverse events, the expanding role of nurse practitioners in US health care, testosterone replacement therapy, HIV treatment, and interventions for numerous chronic conditions. Our use of large administrative databases provide us with a number of methodologic advantages. First, unlike RCTs—which typically require participants to meet restrictive inclusion criteria based on age, race, gender, and comorbid disease—these data sources allow us to examine a wide range of clinically, racially, socioeconomically, and geographically diverse real-world patients. Second, they offer a high degree of statistical power which allows us to examine rare outcomes. Third, patients in these databases have a much longer average duration follow-up than do most RCTs (which typically last <1 year).
This increased follow up duration provides us with a more complete understanding of how a treatment operates across a representative trajectory of a chronic condition. Finally, large databases allow us to examine patients with comorbid diseases and multiple medications, most of whom would be excluded from RCTs. Over the last 15 years, our group—which includes diverse faculty with expertise in epidemiology, biostatistics, and research design—has received multiple NIH grants and has published in a number of high impact journals including the New England Journal of Medicine, JAMA, the Lancet, JAMA Internal Medicine, JNCI, and the American Journal of Medicine.