142nd APHA Annual Meeting and Exposition

Annual Meeting Recordings are now available for purchase

303608
How efficient and inefficient hospitals differ

142nd APHA Annual Meeting and Exposition (November 15 - November 19, 2014): http://www.apha.org/events-and-meetings/annual
Tuesday, November 18, 2014 : 10:50 AM - 11:10 AM

Michael D. Rosko, PhD , Health Care Management, Widener University, Chester, PA
Herbert Wong, PhD , CDOM, Agency for Healthcare Research and Quality, Rockville, MD
Improved hospital cost-efficiency can bend the cost curve without adversely affecting access. Stochastic Frontier Analysis (SFA) was used to estimate hospital cost-efficiency.   Because it was important to control for variations in quality and patient burden of illness, the study was restricted to hospitals in 39 states that supplied administrative data (from which quality indicators were derived) to the Agency for Healthcare Research and Quality (AHRQ) Health Care Cost and Utilization Project (HCUP) during the entire study period. The analytical file included 1,786 hospitals that reported complete data in during the period 2006-10. We used a balanced panel design and merged hospital-specific data from data files compiled by the American Hospital Association and the Center for Medicare and Medicaid Services.  Quality indicators came from the application of the AHRQ Quality Indicator (QI) software to HCUP data.  We used the Battese-Coelli simultaneous SFA model to estimate cost-inefficiency (reversed coded to reflect efficiency to simplify interpretation). The model included standard cost function variables such as admissions, outpatient visits, post-admission patient days, and prices of capital and labor; as well variables for case-mix, service-mix and quality. Inefficiency-effects variables were used to represent a variety of internal and external environment pressures for efficiency including ownership, competition, and payment policies. Statistical tests supported a translog cost function and a truncated-normal distribution for the inefficiency component of the residual.

Preliminary analysis found that the average estimated cost-efficiency was 90.35%. Our analysis focuses on performance measures, characteristics and market variables of the top and bottom 100 hospitals as ranked by cost- efficiency(group means of 97.04% and 70.03%) . We found to the most efficient hospitals tended to have lower average costs ($11,094 vs. $30,610), use fewer employees (0.3342 vs. 0.5873), earned a higher operating margin (0.0979 vs.     -0.2292). The most efficient group of hospitals tended to be non-teaching, for-profit, and members of multi-hospital systems. The least efficient hospitals had more academic medical centers, were publicly owned and much less likely to be a system-member. The highly efficient hospitals relied more on Medicare (45.19% of admissions vs. 38.54%) but less on Medicaid (20.65% vs 23.09%). Not consistent with expectations, hospitals in the most efficient group had a lower mean occupancy rate (59.27% vs. 67.20%)  and were located in areas with a lower HMO penetration rate (18.55% vs 35.48%). The later result may be explained by a greater preference for HMOs in high-cost areas. The paper concludes with policy implications.

Learning Areas:

Biostatistics, economics

Learning Objectives:
Describe how performance varies in very efficient and very inefficient hospitals Explain how stochastic frontier analysis can measure hospital efficiency Discuss implications of correlates of high and low cost-efficiency on public policy

Keyword(s): Hospitals, Economic Analysis

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I am the principal investigator on this project. I have published over 5 papers using stochastic frontier analysis
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.

Back to: 4152.0: Health economics modeling