Mobile Edition
Print
Get the Mag
Weekly UpdatesEXECUTIVE SUMMARY
The primary research question this study addresses is, do size and ownership type make a difference in the efficiency and cost results of hospitals in Washington State? A further question is, what factors might explain such differences? The data source is the hospital financial data reports Washington hospitals submit to the Washington Department of Health. The sample was restricted to not-for-profit and government-owned hospitals, given that these ownership types are predominant in Washington State, and there are only two investor-owned hospitals.
The measures of efficiency and cost represent the generally accepted financial indicators derived from the healthcare financial management literature. Cost and efficiency in these hospitals are analyzed using five efficiency ratios and five cost measures. The results are significant for five of the ten measures studied. Measured by occupancy percentage, small and large not-for-profit hospitals appear to achieve higher efficiency levels than government-owned hospitals do, but the larger hospitals of both ownership types report greater efficiency than that achieved by smaller hospitals. In terms of costs, small, not-for-profit hospitals report comparable costs to those of the largest hospitals, likely because 70 percent of the small not-for-profits are critical access hospitals.
These findings deserve further study on a regional or national level. A more scientific study of the efficiency and cost of hospitals by size and ownership type would be important to control for case mix, scope of services, and payer mix. Such studies can generate important findings about the relationship of hospital size and ownership type to efficiency and cost. Conducted on a national level, such studies would provide policymakers with the empirical data they need to make decisions regarding the types of hospitals to encourage or discourage in the future.
**********
Hospital size has long been an area of discussion and debate in the U.S. healthcare industry. Questions have consistently focused on cost management or efficiency in large versus small hospitals. A persistent question among researchers is whether efficiencies are associated with larger facilities through economies of scale, or if there are alternate scenarios that play a significant part in hospital cost and efficiency.
PRIOR STUDIES
Researchers have used a wide variety of performance measures to compare hospital performance by organization size. In an earlier study, Coyne (1982) examined performance differences between system and independent hospitals using two cost measures (cost per case and payroll per patient day) and two efficiency measures (admissions per bed and full-time equivalents [FTEs] per occupied bed). Griffith, Alexander, and Jelinek (2002) examined cash flow, asset turnover, mortality, complications, length of inpatient stay, cost per case, occupancy, change in occupancy, and percent of revenue from outpatient care. When considering the content validity, reliability, sensitivity, and independence of all nine variables, the authors found that all measures except the two occupancy measures are good gauges of hospital performance. Pink and colleagues (2006), with a technical advisory group, created a financial indicators report specifically for critical access hospitals (CAHs). It includes 20 ratios found to be useful by the chief financial officers of CAHs for measuring profitability, liquidity, revenue, cost, and utilization. Griffith and colleagues (2006) analyzed Medicare data from more than 2,500 hospitals for a five-year period ending in 2003 that showed only a few of their nine measures exhibited signs of improvement, with most indicating volatility or only modest improvements.
Prior hospital performance research findings have been inconclusive in regard to hospital size, such that further study is needed. Yafchak (2000) examined the possibility that hospitals gain economies of scale as size increases. He found that prior to 1994 there were diseconomies of scale in nonteaching hospitals, and that from 1989 to 1997 there were economies of scale, overall, in larger hospitals. Ozcan and Luke (1993) found that hospital ownership and percentage of Medicare were the factors most associated with hospital efficiency, and facility size was consistently and positively related to efficiency due to economies of scale.
This article analyzes the cost and efficiency by size of not-for-profit and government-owned hospitals in the state of Washington. Five efficiency ratios and five cost measures were used. The primary research question is, do size and ownership type make a difference in the efficiency and cost results of hospitals in Washington state? A further question addressed is, what factors might explain the results of this analysis and provide some recommendations for managerial policies and practices in hospitals?
METHODS
Measures and Data
The measures of efficiency and cost represent the generally accepted financial indicators derived from the healthcare financial management literature. The data source is the financial reports hospitals submit to the Washington Department of Health. The sample was restricted to not-for-profit and government-owned hospitals, given that these are the predominant ownership types in Washington State, and there are only two investor-owned hospitals (see Table 1).
The study sample accounts for 98 percent of the hospitals in the state. The study uses three size categories: small (1-40 beds), medium (41-150 beds), and large (151 or more beds). These size categories were chosen because of the relatively even distribution of hospitals across the three. A national data set might benefit from more size categories, particularly for the larger facilities. In Washington State, there are only 27 hospitals with more than 150 beds, 35 small and 34 medium-sized hospitals; further, there is an insufficient sample of facilities in excess of 200 beds. Indeed, in considering statistical power for comparative testing purposes, additional size categories cannot be justified. Given cost-based reimbursement for the small size category of hospitals, special consideration of the CAH is provided in the Discussion section.
The industry averages are represented by the median values for the year 2004 and for the Far West Region, to account for regional variations. As noted in Table 2, the industry averages are derived from the 2006 Almanac of Hospital Financial and Operating Indicators compiled by Ingenix (Parkinson 2006).
Efficiency indicators. The five efficiency indicators in this study are frequently used as measures of hospital performance. It is important to note that the occupancy percentage is based on the available beds and not on the licensed beds (see Table 2).
Cost indicators. The five cost indicators in this study are frequently used as measures of hospital performance. Three of the five cost indicators are adjusted to isolate admissions, discharges, and patient days associated with acute care activity by excluding skilled nursing facility (SNF) and swing beds in the study hospitals.
All measures of cost and efficiency are the mean values for the reporting year 2005. The mean values represent the average for each given size and ownership category.
Histograms were used to examine the distributions of the dependent variables. Two variables (cost per adjusted patient day and cost per adjusted admission) are skewed in their distribution and have been logarithmically transformed to create a more symmetrical distribution and therefore allow a fair statistical test.
Data were tested for differences between hospital sizes (small, medium, and large), for differences between ownership type (not-for-profit versus government-owned), and for differences due to the interaction between hospital size and ownership type using a two-way analysis of variance (ANOVA). SPSS 15.0 for Windows was the statistical package used for conducting the ANOVA tests. Post hoc comparisons of means were examined using Scheffe's method. Results were considered statistically significant when the probability value was less than 5 percent.
RESULTS
The five key results are as follows:
* Current asset turnover results show that size matters but ownership type (by itself) does not in that this measure of efficiency is significantly lower (p < 0.001) in the small hospitals than in the medium and large hospitals for both ownership types.
* Occupancy percentage results show that size and ownership type matter in that this measure of efficiency is highly significant for the main effects of size and ownership type and their interaction.
* Cost per adjusted patient day results show that hospital bed size matters but ownership type does not in that this cost measure is significantly higher in large hospitals than in medium-sized hospitals (Scheffe's p = 0.031).
* FTEs per adjusted patient day results show that size does not matter but ownership type does in that FTEs are higher among government hospitals than among not-for-profit hospitals, irrespective of hospital size, with marginal significance (p = 0.047).
* Salary per FTE results show that hospital size and ownership type matter in that this cost measure is higher in the not-for-profit hospitals than in the government hospitals (p = 0.015) and higher in the larger hospitals than in the small and medium-sized hospitals (p = 0.027).
The two-way ANOVA p-values are presented for bed size, ownership, and the interaction of bed size and ownership (see Table 3). Of the ten ratios that include two efficiency results and three cost results, five are statistically significant, with a probability value less than five percent.
More specifically, two of the five efficiency ratios show significant results, including current asset turnover (Ratio 3) and occupancy percentage (Ratio 4). Further, three of the five cost ratios show significant results, including cost per adjusted patient pay (Ratio 6), FTEs per adjusted patient day (Ratio 9), and salary per FTE (Ratio 10).
FEED