Determinants of Utilization and Expenditures for Episodes of Ambulatory Physical Therapy Among Adults

Abstract

Comprehensive information on determinants and patterns of use and spending for ambulatory physical therapy services is needed to inform health planning and policy decisions. Most research in the literature on this topic is limited to specific payers, age groups, and conditions. The purpose of this study was to examine factors associated with the resource intensity of physical therapy episodes for adults in the United States as measured by number of visits and expenses per visit. This study was a secondary analysis of longitudinal survey data from the Medical Expenditure Panel Survey (MEPS) panels 9, 10, and 11. An analytic file was created based on data from the longitudinal data files for 3 MEPS panels and the annual office-based and hospital outpatient event files. A total of 1,377 episodes of physical therapy care were identified. Variation in both the total number of visits per episode and expenses per visit was examined by fitting regression models to evaluate the effects of selected independent variables classified into 4 categories: episode-level variables, demographic characteristics, geographic variables, and health status indicators. Average total expenses per episode (in 2007 dollars) were $1,184 (median=$651), with an average number of visits per episode of 9.6 (median=6.0) and average expenses per visit of $130 (median=$95). Significant variation by geographic characteristics, sex, and one comorbid condition (high blood pressure) was found in the number of visits model. In the expenditures model, expenses per visit were associated with age/insurance coverage, setting (hospital outpatient versus office based), primary condition category, and mental health status. Limitations Limitations include limited sample sizes of physical therapy users and lack of detailed clinical information. Variability in the resource intensity of physical therapy episodes is influenced to some degree by nonclinical variables.