Driving Precision Health Care through Heterogeneous Outcome Analysis

Operations Management;Health Care;Empirical Econometrics;Machine Learning;Business (General);Business and Economics;Business Administration

Wang, GuihuaLikosky, Donald ;

University of Michigan

关键词: Operations Management;    Health Care;    Empirical Econometrics;    Machine Learning;    Business (General);    Business and Economics;    Business Administration;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/149999/guihuaw_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

This dissertation is based on three essays that examine how to measure the heterogeneity of patient outcomes using readily available data, how to use the results to generate patient-centric outcome information, and how outcome data can be used to benefit patients, payers, and providers. In the first essay, we document a wide variation in quality among 188 surgeons at 35 hospitals in New York State that perform mitral valve surgery. Our analysis shows that patients of different demographics and levels of acuity benefit differently from elite surgeons. We estimate that the total societal benefits from using our proposed patient-centric information are comparable to those achievable by enabling the best surgeons to treat 10%-20% more patients under currently available population-average information. In the second essay, we develop a technique that incorporates the instrumental variable method into a causal tree to correct for potential endogeneity biases in heterogeneous treatment effect analysis using observational data. The resulting instrumental variable tree (IV tree) approach partitions subjects into subgroups with similar treatment effects within subgroups and different treatment effects across subgroups. In the third essay, we provide empirical evidence that outcome differences between health care providers are heterogeneous across different patients. We then use the IV tree approach to identify patient types that exhibit significant differences in outcome quality. After that, we quantify the differences in patient outcomes between providers in a (patient-centric) manner that is useful to individual patients. Lastly, we show that providing patient-centric outcome information not only helps patients choose providers but also helps providers identify areas for improvement and payers design cost-effective payment.