Profit Efficiency and Non-profit Status: A Stochastic Frontier Approach

Objective

To estimate the profitability and profit efficiency differences for common surgical procedures between non-profit and for-profit hospitals.

Data

We use hospital claims (N=13,418) and financial data from 10 geographically similar California hospitals in 2008 and 2010. The claims cover all knee, hip, lumbar fusion, cervical fusion, pacemaker, and stent procedures for each hospital. For each claim, we have information on the reimbursed amount, device cost, drug cost, supply cost, and other costs. We use this information to calculate the profitability of each claim. Half of the hospitals are non-profit and half are for-profit hospitals. Non-profit hospitals account for approximately 70% of the claims. The claims are supplemented with hospital-level administrative and financial data, which we use to control for hospital staffing and overhead costs.

Methods

We use linear regressions to test the profitability differences between non-profit and for-profit hospitals. We use stochastic frontier models to examine the efficiency of non-profit and for-profit hospitals. Claim-level profits were normalized by dividing profits by total input costs. We control for patient characteristics (age, insurance type, and condition severity), hospital volume, other inputs in each hospital’s cost function (number of full-time employees, number of surgeons, number of beds, square feet devoted to surgery services), and year fixed effects.

Results

Across all procedures, non-profit hospitals have lower normalized profits (-0.263 (0.136)). This effect is driven by lower normalized profits for cardiovascular procedures (-0.353 (0.112)). The difference in profits appears to be driven by a larger share of claims with negative profits for non-profits than for-profits. Non-profit status is associated with a lower efficiency frontier for all procedure types. In addition, for profit hospitals have higher predicted profit efficiency of 0.05 (95% CI: 0.04-0.06) percentage points for all procedures 0.03 (95% CI: 0.02-0.04) percentage points for orthopedic procedures, and 0.02 (95% CI: 0.004-0.029) percentage points for cardiovascular procedures. For both hospital types, mean predicted profit efficiency ranged from 0.89 (95% CI: 0.88-0.89) for both procedure types to 0.84 (95% CI: 0.83-0.84) for cardiovascular procedures.

Conclusion

In this setting, non-profit hospitals have slightly lower profitability and lower profit efficiency for surgical device procedures than for-profit hospitals.