Evaluation of Bundled Payments for Care Improvement Model 1

Official Title
Evaluation and Monitoring of the Bundled Payments for Care Improvement Model 1 Initiative
Baltimore, Maryland
5/2013 - 12/2016
IMPAQ Health
Centers for Medicare and Medicaid Services
Sub To
Applied Econometrics
Data Analytics
Data Collection, Analysis & Reporting
Quantitative Data Analysis
Quasi-experimental Impact Evaluations
Survey Analysis
Survey Design and Sampling
Survey Methodology

IMPAQ’s work supported the Center for Medicare & Medicaid Innovation, by monitoring and evaluating their Bundled Payment for Care Improvement (BPCI) Model 1 initiative. BPCI Model 1 allows gainsharing between participating hospitals and physicians when internal hospital costs fall below a discounted Medicare Severity-Diagnosis Related Group (MS-DRG) payment and total Medicare payments during the admission through 30-day post-discharge period fall below a benchmark. The goal is to provide an incentive for hospitals and physicians to coordinate care during inpatient episodes, to improve quality and efficiency, and to reduce costs.

We used a difference-in-differences approach to estimate the impact of the program. First, we identified comparison hospitals using Mahalanobis distance matching. During the two base years and in the first option year, we led the design, implementation, and analysis of a beneficiary survey called the Patient Health and Experience Survey (PHES). The PHES collected data on functional status, pain, care coordination, and access to care for patients discharged from hospitals participating in BPCI Model 1. Our analysis examined how care changed in response to incentives introduced by the BPCI Model 1.

This work builds on our history of Medicare claims analysis for implementation and evaluation work while taking advantage of our subject matter expertise in innovative healthcare payment models. Our experienced managing director and principal research economist, project manager, and researchers bring a range of experiences, which includes developing large, claims-based analytic data sets and understanding quasi-experimental methods and their applications to evaluation and policy analysis.