PhD Defense by Andrew ElHabr

Primary tabs

Thesis Title: Essays on Healthcare Outcomes and Value-based Care



Dr. Turgay Ayer, School of Industrial and Systems Engineering, Georgia Institute of Technology


Thesis Committee:

Dr. Gian-Gabriel Garcia, School of Industrial and Systems Engineering, Georgia Institute of Technology

Dr. Lauren Steimle, School of Industrial and Systems Engineering, Georgia Institute of Technology

Dr. Benjamin Haaland, School of Medicine, University of Utah

Dr. Judy Gichoya, School of Medicine, Emory University


Date and Time: Tuesday, August 9th, 2022, 2:00 pm EDT

Location: Groseclose 402

Meeting Link: Zoom



Value-based healthcare is the general healthcare delivery framework in which service quality is prioritized over quantity. The implementation of this framework ranges from insurance payment design to clinical decision-making based on best practices established through analyses of costs and outcomes. This thesis utilizes a variety of observational data, including data from several large administrative claims databases, registries, and electronic medical records, and applies principles of value-based care to make advances in various areas, such as pharmaceutical contracts, severity adjustment for outcome studies of acute ischemic stroke patients, and appropriate use of diagnostic imaging and blood products in emergency care settings.


In Chapter 2, we analyze a national claims database of commercially insured and Medicare Advantage enrollees to evaluate the emergency department utilization of advanced neuroimaging in the last decade and a half. We observe that the substantial rise of age-adjusted utilization of advanced neuroimaging in these populations mirrors that in the Medicare fee-for-service population.


In Chapter 3, we use a game-theoretic stylistic model to study under what market conditions and drug characteristics insurers and pharmaceutical manufacturers are better off engaging in an outcomes-based contract over a nominal pricing contract when there are two distinct target patient groups for the drug. Outcome-based contracts, as opposed to nominal pricing contracts, link payments for drugs to drug effectiveness. We also investigate the impact that the equilibrium pricing and cost-sharing decisions have on patient welfare.


In Chapter 4, we first establish a surrogate outcome for long-term functional ability after an acute ischemic stroke incident using registry data from a comprehensive stroke center. Then, we leverage two large claims databases to build an interpretable model to impute a missing clinically-relevant severity measure using features based on established risk adjustment categories, and then compare model performance to various baselines. We also contemplate the implications of developing risk adjustment models like this based on only one database.


In Chapter 5, we analyze emergency hemorrhage management in trauma patients under a variety of scopes. First, at the national level, we investigate racial disparities in adjusted utilization rates and outcomes of non-operative management, operative management, and interventional radiology (IR) procedures in patients suffering isolated pelvic, splenic, liver, kidney, and poly-trauma injuries. Secondly, we estimate the impact of diagnostic imaging prior to IR procedures on mortality. Lastly, we study associations between mortality and 24-hour trends in vitals and rates of blood product given in ultra-massive transfusion patients from a local level I trauma center. 


  • Workflow Status: Published
  • Created By: Tatianna Richardson
  • Created: 07/28/2022
  • Modified By: Tatianna Richardson
  • Modified: 07/28/2022