Regression and Data Envelopment Analysis Methods to Assess Practice Efficiency
Chapter
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
This study used regression and data envelopment analysis (DEA)to estimate efficiency and relative efficiency of medical group practices in the United States. Three years of data from the Medical Group Management Association (MGMA) practice cost survey were analyzed. MGMA surveyed practices to determine revenues, costs, labor and capital, visits/encounters, and relative value units (RVUs). Thus, adequate information to determine relationships between inputs and outputs was available. Our approach included an initial descriptive analysis of primary care practice inputs, outputs, and revenues. Data for three consecutive survey years was then combined to form a panel dataset for the subset of practices that participated in the survey for multiple years. From the input, output, and revenue data, we determined efficiency frontiers from DEA. DEA identified the efficiency frontier and then identified the relative efficiency of medical practices in the data set. Use of both methods can inform practice efficiency improvement strategies. The regression models included multiple dependent variables (i.e., net operating costs per procedure inside the practice and RVUs). Models examining the panel data used a medical cost inflation adjustment. Coefficients for the various regression parameters (e.g., rural location, number of physician-extenders, total staff in the practice, etc.) were interpreted to show the marginal impact on the practice efficiency measure. Finally, we compared findings and interpretation of the models and discuss utility of the two approaches in understanding primary care practice efficiency.
