The measurement of productivity in an emergency medicine group practice

Abstract

Study objectives: Productivity in a medical service organization is difficult to define, and its measurement is a controversial issue. Relative value units (RVU) have been used by some organizations, but this index takes into account only work product. A new emergency medicine financial productivity index (EMPI) has been developed for testing and validation. We calculate the EMPI defined as revenue per patient service hour per work hour and determine the correlation of the EMPI with RVU. Methods: This was a retrospective data review of fee-for-service emergency medicine group practice comprising 16 sites in the northern New Jersey region. A total of 210 physicians working at 16 emergency departments were studied for each month from January 2003 through March 2004, accounting for 825,549 consecutive patient encounters. Input variables included hours worked, charges per patient, collections per patient, length of stay (LOS) per patient, RVUs per patient encounter, and volume of patients. RVU work units were compiled using federal 2003 guidelines. The EMPI was defined as the revenue per patient-hour per work-hour and is derived as follows: EMPI = {{Average revenue per patient encounter ÷ Average LOS per patient encounter} ÷ Hours worked} × Patient volume The EMPI calculation was performed 3 ways where the average revenue per patient encounter was charge based, collection based, or the RVU work product index. The financial EMPI (both charge and collection based) by physician and hospital site was derived and plotted against the corresponding RVU function to determine correlation. The EMPI was also derived by sample disease states, including ankle sprain, asthma, upper respiratory infection (URI), and chest pain. The EMPI by sample disease state was plotted against the RVU function to determine correlation. Results: The charge-based financial EMPI by physician and hospital site correlated with the RVU function for all patient encounters by month at all ED sites and yielded a remarkably linear relation (R=0.991 charge based by physician). The collection-based financial EMPI also correlated with the RVU function, but not as well as the charge-based calculation (R=0.877). The correlation of the charge-based financial EMPI by disease states with the RVU function was also very linear (R=0.993 for asthma, R=0.996 for ankle sprain, R=0.978 for URI, and R=0.991 for chest pain). The collection-based financial EMPI was also linear but had less of a correlation with the RVU function (R=0.882 for asthma, R=0.886 for ankle sprain, R=0.967 for URI, and R=0.908 for chest pain). Conclusion: A productivity index for emergency medicine practice has been developed that incorporates revenue (charges versus collections), hours worked, LOS, and patient volume. This model is financially based and does not incorporate other physician-centric attributes such as patient satisfaction, hospital committee participation, or research publications. This index is shown to correlate well with RVU calculations. The correlation appears to occur by physician, ED site, and disease state for all patient encounters. Further studies are warranted for validation of this model.