Confronting the Statistical Challenges of Measuring Health Care Provider Performance

• Activity Number: 143
• Type: Invited
• Date/Time: Monday, July 30, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Rebecca Hubbard*+ and Victoria Ding and Carolyn Rutter
• Companies: Group Health Research Institute and University of Washington and Group Health Research Institute
• Keywords: provider profiling ; mental health ; classification

Abstract

Provider profiling as a means to describe and compare performance of health care professionals has gained great momentum recently. While provider profiling traditionally focused on the identification of outliers, it increasingly has been used to classify providers as top- or bottom-performers, regardless of whether their performance lies outside the norm. For instance, Group Health of Washington state (GH) initiated an aggressive incentive program that awards bonuses to community mental health care practitioners contracting with Group Health who score above the 90th percentile on quarterly provider profiles. Motivated by the context of mental health care provider performance at GH, we used a simulation study to compare the performance of commonly used methods for identifying high performing providers, evaluating classification performance using sensitivity and specificity. We will present results from our simulation study and demonstrate the performance of these methods using data for mental health care providers from GH. Based on our findings, we caution against the use of any of the existing classification methods in settings of low between-provider variability.