Online Program Home
My Program

Abstract Details

Activity Number: 197 - SPEED: Government and Health Policy
Type: Contributed
Date/Time: Monday, July 30, 2018 : 10:30 AM to 11:15 AM
Sponsor: Health Policy Statistics Section
Abstract #332907
Title: New Applications of Machine Learning to Estimating Large Physician Demand Models
Author(s): Bryan Sayer* and William Encinosa
Companies: Social & Scientific Systems, Inc. and Agency for Health Care Quality and Research
Keywords: Machine learning; semiparametric estimation
Abstract:

One of the largest recent transformations in healthcare has been the massive consolidation of physician groups. Little policy work has been done on this due to the lack of physician demand models, which differ from hospital demand models due to their uncommonly large choice sets of docs. Here we develop new methods for estimating physician discrete choice models in Medicare. We compare estimation methods: parametric, semiparametric, and machine learning methods. We also explore methods of randomly reducing the choice sets, similar to McFadden 1978.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program