Healthcare operations models tend to assume clean data, stable behavior, and perfect compliance. Real clinics and hospitals don't work that way. I study how to design scheduling, staffing, and workflow systems that are mathematically sound — and behaviorally survivable.
The simplest operational model that survives real human behavior. Not simple for its own sake — simple because transparency, trust, and recalibratability are what real teams actually adopt and sustain. The proof case: across 17,314 surgical cases, a parsimonious surgeon–procedure model outperformed a commercial machine-learning scheduler by 13–29 percentage points and cut the median prediction error in half. See the working papers →
My research philosophy is rooted in a conviction: mathematical optimization in healthcare operations is necessary, but insufficient on its own. True efficiency cannot exist in a behavioral vacuum. For decades, operations research has treated clinical environments as frictionless systems — generating highly optimized models that frequently fail upon implementation due to cognitive load, misaligned incentives, and human resistance.
I view healthcare delivery — whether in large hospital networks, specialized ambulatory centers, or specialized longevity and aesthetics clinics — as complex socio-technical systems. My overarching goal is to define and operationalize contextual parsimony: the pursuit of the simplest, most efficient operational model that survives the messy reality of human behavior.
Integrating change management, leadership, and behavioral economics into traditional operations research is, I believe, the only way to design systems that clinicians will naturally adopt — and that will sustainably improve performance and patient outcomes. My work replaces the assumption of frictionless compliance with a more realistic theoretical framework of system performance in high-variability, high-stress environments.
Highly optimized schedules fail because they demand perfect compliance from highly stressed people. This stream investigates the behavioral design of scheduling and workflow architecture — choice architecture and EHR defaults that guide clinicians toward efficient choices without mandates. Target environments: anesthesia delivery, surgical block scheduling, specialized outpatient procedures.
Operational efficiency degrades as initial compliance fades. This stream examines how financial, social, and structural incentives interact with operational models to sustain — or erode — designed capacity.
Even the most behaviorally sound system requires implementation. This stream bridges operations with organizational theory: the role of clinical leadership in the adoption of optimized operational change.
Books, dissertation research, and working papers — each listed with its honest status. Every claim traces to a source.
A three-arm pragmatic cluster-randomized trial across approximately 114 outpatient clinics in a rural integrated health system, testing whether a structured execution-support workflow preserves appointment capacity beyond reminder-rich usual care. Three co-primary outcomes: no-show rate, cancellation rate, and slot-utilization rate. The conceptual contribution — slot stewardship — redefines scheduling success as capacity preservation rather than raw attendance. Two empirical papers planned. PhD in Business, Virginia Tech.
Working-paper titles and findings are stated as of manuscript stage and may change in peer review. New submissions, acceptances, and preprints are announced in the newsletter.
I'm actively seeking collaborators on behavioral operations in clinical settings — scheduling and no-show interventions, choice architecture in EHRs, performance-management field experiments, and change-management measurement. I bring what most academic partners can't: live clinical environments, proprietary operational data, and a 114-clinic field-trial infrastructure.
Faculty, health systems, and doctoral researchers welcome. If you have a question that needs a real clinical setting — or a setting that needs a rigorous question — let's talk.