Frontier estimation: how useful a tool for health economics?

Electronic medical records and cost efficiency in hospital medical-surgical units

This study examines the impact of electronic medical records (EMRs) on cost efficiency in hospital medical-surgical units. Using panel data on California hospitals from 1998 to 2007, we employed stochastic frontier analysis (SFA) to estimate the relationships between EMR implementation and the cost inefficiency of medical-surgical units. We categorized EMR implementation into three stages based on the level of sophistication. We also examined the effects of specific EMR systems on cost inefficiency. Our SFA models addressed potential bias from unobserved heterogeneity and heteroskedasticity. EMR Stages 1 and 2, nursing documentation, electronic medication administration records, and clinical decision support were associated with significantly higher inefficiency.

What have we learned from the application of stochastic frontier analysis to U.S. hospitals?

This article focuses on the lessons learned from stochastic frontier analysis studies of U.S. hospitals, of which at least 27 have been published. A brief discussion of frontier techniques is provided, but a technical review of the literature is not included because overviews of estimation issues have been published recently. The primary focus is on the correlates of hospital inefficiency. In addition to examining the association of market pressures and hospital inefficiency, the authors also examined the relationship between inefficiency and hospital behavior (e.g., hospital exits) and inefficiency and other measures of hospital performance (e.g., outcome measures of quality). The authors found that consensus is emerging on the relationship of some factors to hospital efficiency; however, further research is needed to better understand others. The application of stochastic frontier analysis to specific policy issues is in its infancy; however, the methodology holds promise for being useful in certain contexts.

Principal agent relationships and the efficiency of hospitals

The efficiency of hospitals is an important political issue and has been the subject of a number of studies. Most studies find evidence for inefficiency but provide no theoretical explanations for differences in efficiency. This study used principal agent theory to explain differences in efficiency between hospitals. Two agency issues are examined: (1) quality of care in the relationship between hospital and patient, and (2) internal organisation, i.e. the relationship between the hospital and its main departments. It was found that efficiency and quality go together. This implies that the potential harmful information asymmetry between hospitals and patients does not appear to be a major problem, because increasing efficiency does not seem to reduce quality. Further, we find no relationship between the efficiency of departments and the efficiency of the entire hospital. The interest of hospital departments is currently not in line with the interests of the entire hospital.

What factors affect the productivity and efficiency of physician practices?

BACKGROUND

Increasing the productivity and efficiency of physician practices could help relieve the rapid growth of US healthcare costs and the expected physician shortage. Radiology practices are an attractive specific focus for research on practices' productivity and efficiency because they are home to many purportedly productivity-enhancing operational technologies. This affords an opportunity to study the effect of production technology on physicians' output. As well, radiology is a leader in the general movement of physicians out of very small practices. And imaging is by the fastest-growing category of physician expenditure.

RESEARCH DESIGN

Using data from 2003 to 2007 surveys of radiologists, we estimate a stochastic frontier model to study the effects of practice characteristics, such as work hours, practice size, and output mix, and technologies used in work production, on practices' productivity and efficiency.

RESULTS

At the mean, the elasticities of output with respect to practice size and annual hours worked per full-time physician were 0.73 and 0.51, respectively. Some production technologies increase productivity by 15% to 20%; others generate no increase. Using "nighthawks"--ie, contracting out after-hours work to external firms that consolidate workflow--significantly increases practice efficiency.

CONCLUSIONS

The general US trend toward larger practice size is unlikely to relieve cost or physician shortage pressures. The actual effect of purportedly productivity-enhancing operational technologies needs to be carefully evaluated before they are widely adopted. As the recently-developed innovations of nighthawks and hospitalists show, practices should give more attention to a possible choice to "buy," rather than "make," part of their output.

Efficiency and Financing in Catering Primary Health Care to the Poors

Public health care system is a crucial element of social welfare, particularly in underdeveloped areas where a majority of the populace lingers in abject poverty and deprivation. The system is typically assumed to be cheap and based on a public welfare perspective. However, it is argued that in the name of reform process, it is usually targeted as a soft option. Using a panel data of all the (reported) hospitals in the Indian state of West Bengal over the seven-year period, 1997–2003, we try to answer this question. The framework of input–output data is used in order to assess the performance of these hospitals. The non-parametric Data Envelopment Analysis (DEA) is used to assess the efficiency of these hospitals over these periods. The results do suggest initial adverse impact of the faulty reform process but indicate some later revival. Huge intra-state differences persist. Surprisingly, economically advanced regions are adjudged, too, more efficient than their less developed counterpart. Public expenditure also appeared to be positively correlated with efficiency. In fact, our study indicates that one of the most important factors in determining efficiency is per hospital real health expenditure.

Measurement of efficiency of the dialysis sector in Turkey using data envelopment analysis

OBJECTIVES

Limited prior work exists on the efficiency of the dialysis sector characterized by steady increase in the dialysis population, providers, and dialysis expenditures in Turkey. The study aimed to examine technical efficiency among producers of dialysis treatment in Turkey.

METHODS

Cross-sectional data were used from the 2008 Turkish Statistical Yearbook of Dialysis for total 830 dialysis facilities. Efficiency was measured using data envelopment analysis technique.

RESULTS

Results drawn from an input-oriented variable-returns-to-scale model showed technical efficiency from only 3% of facilities, as well as 50% reduction in inputs. Mean efficiency score was found higher among facilities that were freestanding, private, affiliated with international chains, older, and located in Istanbul.

CONCLUSION

Efficiency enhancement in dialysis production in Turkey should be placed on the priority agenda, along with careful evaluation of impact of the already-inserted changes on efficiency.

A systematic review of health care efficiency measures

OBJECTIVE

To review and characterize existing health care efficiency measures in order to facilitate a common understanding about the adequacy of these methods.

DATA SOURCES

Review of the MedLine and EconLit databases for articles published from 1990 to 2008, as well as search of the "gray" literature for additional measures developed by private organizations.

STUDY DESIGN

We performed a systematic review for existing efficiency measures. We classified the efficiency measures by perspective, outputs, inputs, methods used, and reporting of scientific soundness.

PRINCIPAL FINDINGS

We identified 265 measures in the peer-reviewed literature and eight measures in the gray literature, with little overlap between the two sets of measures. Almost all of the measures did not explicitly consider the quality of care. Thus, if quality varies substantially across groups, which is likely in some cases, the measures reflect only the costs of care, not efficiency. Evidence on the measures' scientific soundness was mostly lacking: evidence on reliability or validity was reported for six measures (2.3 percent) and sensitivity analyses were reported for 67 measures (25.3 percent).

CONCLUSIONS

Efficiency measures have been subjected to few rigorous evaluations of reliability and validity, and methods of accounting for quality of care in efficiency measurement are not well developed at this time. Use of these measures without greater understanding of these issues is likely to engender resistance from providers and could lead to unintended consequences.

Testing the hospital value proposition: an empirical analysis of efficiency and quality

PURPOSES

To assess the relationship between hospitals' X-inefficiency levels and overall care quality based on the National Quality Forum's 27 safe practices score and to improve the analytic strategy for assessing X-inefficiency.

METHODOLOGY

The 2005 versions of the American Hospital Association and Leapfrog Group's annual surveys were the basis of the study. Additional case mix indices and market variables were drawn from the Centers for Medicare and Medicaid Services data sources and the Area Resource File. Data envelopment analysis was used to determine hospitals' X-inefficiency scores relative to their market-level competitors. Regression was used to assess the relationship between X-inefficiency and quality, controlling for organizational and market characteristics. Expenses (total and labor expenditures), case-mix-adjusted admissions, length of stay, and licensed beds defined the X-inefficiency function. The overall National Quality Forum's safe practice score, health maintenance organization penetration, market share, and teaching status served as independent control variables in the regression.

FINDINGS

The National Quality Forum's safe practice scores are significantly and positively correlated to hospital X-inefficiency levels (beta = .105, p < or = .05). The analysis of the value proposition had very good explanatory power (adjusted R(2) = .414; p < or = .001; df = 7, 265). Contrary to earlier findings, health maintenance organization penetration and being a teaching hospital were positively related to X-inefficiency. Similar with others' findings, greater market share and for-profit ownership were negatively associated with X-inefficiency.

PRACTICE IMPLICATIONS

Measurement of overall hospital quality is improving but can still be made better. Nevertheless, the National Quality Forum's measure is significantly related to efficiency and could be used to create differential pay-for-performance programs. A market-segmented analytic strategy for studying hospitals' efficiency yields results with a high degree of explanatory power.

The measurement of efficiency and productivity of health care delivery

The measurement of efficiency and productivity of health service delivery has become a small industry. This is a review of 317 published papers on frontier efficiency measurement. The techniques used are mainly based on non-parametric data envelopment analysis, but there is increasing use of parametric techniques, such as stochastic frontier analysis. Applications to hospitals and other health care organizations and areas are reviewed and summarised, and some meta-type analysis undertaken. Cautious conclusions are that public provision may be potentially more efficient than private, in certain settings. The paper also considers conceptualizations of efficiency, and points to dangers and opportunities in generating such information. Finally, some criteria for assessing the use and usefulness of efficiency studies are established, with a view to helping both researchers and those assessing whether or not to act upon published results.

Specialty and full-service hospitals: a comparative cost analysis

OBJECTIVE

To compare the costs of physician-owned cardiac, orthopedic, and surgical single specialty hospitals with those of full-service hospital competitors.

DATA SOURCES

The primary data sources are the Medicare Cost Reports for 1998-2004 and hospital inpatient discharge data for three of the states where single specialty hospitals are most prevalent, Texas, California, and Arizona. The latter were obtained from the Texas Department of State Health Services, the California Office of Statewide Health Planning and Development, and the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project. Additional data comes from the American Hospital Association Annual Survey Database.

STUDY DESIGN

We identified all physician-owned cardiac, orthopedic, and surgical specialty hospitals in these three states as well as all full-service acute care hospitals serving the same market areas, defined using Dartmouth Hospital Referral Regions. We estimated a hospital cost function using stochastic frontier regression analysis, and generated hospital specific inefficiency measures. Application of t-tests of significance compared the inefficiency measures of specialty hospitals with those of full-service hospitals to make general comparisons between these classes of hospitals.

PRINCIPAL FINDINGS

Results do not provide evidence that specialty hospitals are more efficient than the full-service hospitals with whom they compete. In particular, orthopedic and surgical specialty hospitals appear to have significantly higher levels of cost inefficiency. Cardiac hospitals, however, do not appear to be different from competitors in this respect.

CONCLUSIONS

Policymakers should not embrace the assumption that physician-owned specialty hospitals produce patient care more efficiently than their full-service hospital competitors.

Modeling efficiency at the process level: an examination of the care planning process in nursing homes

OBJECTIVE

To examine the efficiency of the care planning process in nursing homes.

METHODS

We collected detailed primary data about the care planning process for a stratified random sample of 107 nursing homes from Kansas and Missouri. We used these data to calculate the average direct cost per care plan and used data on selected deficiencies from the Online Survey Certification and Reporting System to measure the quality of care planning. We then analyzed the efficiency of the assessment process using corrected ordinary least squares (COLS) and data envelopment analysis (DEA).

RESULTS

Both approaches suggested that there was considerable inefficiency in the care planning process. The average COLS score was 0.43; the average DEA score was 0.48. The correlation between the two sets of scores was quite high, and there was no indication that lower costs resulted in lower quality. For-profit facilities were significantly more efficient than not-for-profit facilities.

CONCLUSIONS

Multiple studies of nursing homes have found evidence of inefficiency, but virtually all have had measurement problems that raise questions about the results. This analysis, which focuses on a process with much simpler measurement issues, finds evidence of inefficiency that is largely consistent with earlier studies. Making nursing homes more efficient merits closer attention as a strategy for improving care. Increasing efficiency by adopting well-designed, reliable processes can simultaneously reduce costs and improve quality.

Hospital quality, efficiency, and input slack differentials

OBJECTIVE

To use an advance in data envelopment analysis (DEA) called congestion analysis to assess the trade-offs between quality and efficiency in U.S. hospitals.

STUDY SETTING

Urban U.S. hospitals in 34 states operating in 2004.

STUDY DESIGN AND DATA COLLECTION

Input and output data from 1,377 urban hospitals were taken from the American Hospital Association Annual Survey and the Medicare Cost Reports. Nurse-sensitive measures of quality came from the application of the Patient Safety Indicator (PSI) module of the Agency for Healthcare Research and Quality (AHRQ) Quality Indicator software to State Inpatient Databases (SID) provided by the Healthcare Cost and Utilization Project (HCUP).

DATA ANALYSIS

In the first step of the study, hospitals' relative output-based efficiency was determined in order to obtain a measure of congestion (i.e., the productivity loss due to the occurrence of patient safety events). The outputs were adjusted to account for this productivity loss, and a second DEA was performed to obtain input slack values. Differences in slack values between unadjusted and adjusted outputs were used to measure either relative inefficiency or a need for quality improvement.

PRINCIPAL FINDINGS

Overall, the hospitals in our sample could increase the total amount of outputs produced by an average of 26 percent by eliminating inefficiency. About 3 percent of this inefficiency can be attributed to congestion. Analysis of subsamples showed that teaching hospitals experienced no congestion loss. We found that quality of care could be improved by increasing the number of labor inputs in low-quality hospitals, whereas high-quality hospitals tended to have slack on personnel.

CONCLUSIONS

Results suggest that reallocation of resources could increase the relative quality among hospitals in our sample. Further, higher quality in some dimensions of care need not be achieved as a result of higher costs or through reduced access to health care.

The quantile regression approach to efficiency measurement: insights from Monte Carlo simulations

In the health economics literature there is an ongoing debate over approaches used to estimate the efficiency of health systems at various levels, from the level of the individual hospital - or nursing home - up to that of the health system as a whole. The two most widely used approaches to evaluating the efficiency with which various units deliver care are non-parametric data envelopment analysis (DEA) and parametric stochastic frontier analysis (SFA). Productivity researchers tend to have very strong preferences over which methodology to use for efficiency estimation. In this paper, we use Monte Carlo simulation to compare the performance of DEA and SFA in terms of their ability to accurately estimate efficiency. We also evaluate quantile regression as a potential alternative approach. A Cobb-Douglas production function, random error terms and a technical inefficiency term with different distributions are used to calculate the observed output. The results, based on these experiments, suggest that neither DEA nor SFA can be regarded as clearly dominant, and that, depending on the quantile estimated, the quantile regression approach may be a useful addition to the armamentarium of methods for estimating technical efficiency.

Cost inefficiency and hospital health outcomes

This study explores the association between cost inefficiency and health outcomes in a national sample of acute-care hospitals in the US over the period 1999-2001, with health outcomes being measured by both mortality and complications rates. The empirical analysis examines health outcomes as a function of cost inefficiency and other determinants of outcomes, using stochastic frontier analysis to obtain hospital cost inefficiency scores. The results showed no systematic pattern of association between cost inefficiency and hospital health outcomes; the basic results were unchanged regardless of whether cost inefficiency was measured with or without using instrumental variables. The analysis also indicated, however, that the association between cost inefficiency and health outcomes may vary substantially across geographical regions. The study highlights the importance of distinguishing between 'good' costs that reflect the efficient use of resources and 'bad' costs that stem from waste and other forms of inefficiency. In particular, the study's results suggest that hospital programs focused on reducing cost inefficiency are unlikely to be associated with worsened hospital-level mortality or complications rates, while, on the other hand, across-the-board reductions in cost could well have adverse consequences on health outcomes by reducing efficient as well as inefficient costs.

Effects of ownership, subsidization and teaching activities on hospital costs in Switzerland

This paper explores the cost structure of Swiss hospitals, focusing on differences due to teaching activities and those related to ownership and subsidization types. A stochastic total cost frontier with a Cobb-Douglas functional form has been estimated for a panel of 148 general hospitals over the six-year period from 1998 to 2003. Inpatient cases adjusted by DRG cost weights and ambulatory revenues are considered as two separate outputs. The adopted econometric specification allows for unobserved heterogeneity across hospitals. The results suggest that teaching activities are an important cost-driving factor and hospitals that have a broader range of specialization are relatively more costly. The excess costs of university hospitals can be explained by more extensive teaching activities as well as the relative complexity of the offered medical treatments from a teaching point of view. However, even after controlling for such differences university hospitals have shown a relatively low cost-efficiency especially in the first two or three years of the sample period. The analysis does not provide any evidence of significant efficiency differences across ownership/subsidy categories.

Hospital inefficiency: what is the impact of membership in different types of systems?

The primary objective of this study is to assess whether systematic differences in inefficiency are associated with hospital membership in different types of systems. We employed the Battese/Coelli simultaneous stochastic frontier analysis (SFA) technique to estimate hospital cost inefficiency. Mean estimated inefficiency was 8.42%. Membership in different types of systems was related to estimated cost inefficiency (p < .05). Compared to hospitals that were members of centralized health systems, membership in centralized physician/insurance or decentralized systems was associated with decreased inefficiency; membership in independent systems was associated with increased inefficiency.

Stochastic frontier analysis of hospital inefficiency: a review of empirical issues and an assessment of robustness

Twenty stochastic frontier analysis (SFA) studies of hospital inefficiency in the United States were analyzed. Results from best-practice methods were compared against previously used methods in hospital studies to ascertain the robustness of SFA in estimating cost inefficiency. To compare past studies and analyze new data, SFA methods were varied by (a) the assumptions of the structure of costs and distribution of the error term, (b) inclusion of quality and product descriptor measures, and (c) use of simultaneous and two-stage estimation techniques. SFA results were relatively insensitive to several model variations.

Productivity and employees' organizational justice perceptions in long-term care for the elderly

We examined the associations between productivity, employer characteristics, and context variables, and the organizational justice perceptions of 330 female employees in long-term institutional elderly people care. The productivity measure used was the proportion of the inpatient days to total costs. Employees working in high productivity units experienced higher procedural justice than those working in low productivity units. Hostile employees experienced both the procedures and management as less fair than non-hostile employees. Unit size and resident turnover were negatively and registered nurses percentage positively associated with procedural justice perceptions.

Managed care and the scale efficiency of US hospitals

Managed care penetration has been partly responsible for slowing down increases in health care costs in recent years. This study uses a 1992-1996 Health Care Utilization Project sample of hospitals to analyze the relationship between managed care penetration in local insurance markets and hospital scale efficiency. After controlling for hospital and market area variables, we find that managed care insurance, particularly the preferred provider type, is associated with increases in hospital scale efficiency in tertiary cases. The results presented here are consistent with the view that managed care can lead to reductions in health cost inflation by controlling the diffusion of technology via improvements in the scale efficiency of hospitals.

Addressing the issues that arise in analysing multicentre cost data, with application to a multinational study

Differences in the mean, spread and skewness of cost data collected from different countries present problems for analysis and interpretation. Here we develop generalised linear multilevel models to estimate the effects of patient and national characteristics on costs. Using gamma distributions and multiplicative effects for patient characteristics fitted the data better than models which assumed normal distributions or estimated additive effects. A multilevel gamma model is employed to allow for heterogeneity in the effects of patient case-mix across centres. Analysis of multinational cost data must recognise differences in mean, spread and skewness across centres, as well as the data's hierarchical structure.

Performance-based budgeting in the public sector: an illustration from the VA health care system

This paper estimates frontier cost functions for US Department of Veterans Affairs (VA) hospitals in FY2000 that are consistent with economic theory and explicitly account for cost differences across patients' risk, level of access to care, quality of care, and hospital-specific characteristics. Results indicate that on average VA hospitals in FY2000 operate at efficiency levels of 94%, as compared to previous studies on US private sector hospitals that average closer to 90% efficient. Using these cost frontiers, management systems potentially could be implemented to enhance the equitable allocation of the VA medical care global budget and systematically distribute funds across hospitals and networks. The paper also provides recommendations to improve the efficiency of delivering health care services applicable to public sector organizations.

RUG-III for exploring the association between staffing levels and cost-efficiency in nursing facility care in Finland

The current health policy of the Finnish government with regard to the production of health services strongly emphasizes productivity and efficiency. On the other hand, the staff generally demand better staffing for nursing facility care. This study incorporates the Resource Utilization Groups (RUG-III) in the calculation of the staffing levels and cost-efficiency and explores the association between cost-efficiency and staff levels in nursing facility care in Finland. The results indicate an improvement potential in case-mix-based staffing. Existing labor resources can be used more fairly and effectively while cost-efficiency differences can be diminished by applying the RUG-III for the allocation of the staff.

Estimating technical efficiency in the hospital sector with panel data: a comparison of parametric and non-parametric techniques

BACKGROUND

Policy makers are increasingly interested in developing performance indicators that measure hospital efficiency. These indicators may give the purchasers of health services an additional regulatory tool to contain health expenditure.

OBJECTIVE

Using panel data, this study compares different parametric (econometric) and non-parametric (linear programming) techniques for the measurement of a hospital's technical efficiency.

METHOD

This comparison was made using a sample of 17 Italian hospitals in the years 1996-9.

RESULTS

Highest correlations are found in the efficiency scores between the non-parametric data envelopment analysis under the constant returns to scale assumption (DEA-CRS) and several parametric models. Correlation reduces markedly when using more flexible non-parametric specifications such as data envelopment analysis under the variable returns to scale assumption (DEA-VRS) and the free disposal hull (FDH) model. Correlation also generally reduces when moving from one output to two-output specifications.

CONCLUSIONS

This analysis suggests that there is scope for developing performance indicators at hospital level using panel data, but it is important that extensive sensitivity analysis is carried out if purchasers wish to make use of these indicators in practice.

The cost efficiency and clinical quality of institutional long-term care for the elderly

This study applied the stochastic frontier cost function with inefficiency effects to estimate the association between quality of care and cost efficiency in institutional long-term care wards for the elderly in Finland. We used several clinical quality indicators for indicating adverse care processes and outcomes, based on the Resident Assessment Instrument (RAI)/Minimum Data Set (MDS). Average cost inefficiency among the wards was 22%. We found an association between the clinical quality indicators and cost inefficiency. Higher prevalence of pressure ulcers was associated with higher costs, whereas the higher prevalence of use of depressants and hypnotics increased inefficiency.

Measuring the productive efficiency and clinical quality of institutional long-term care for the elderly

The authors consider the association between productive efficiency and clinical quality in institutional long-term care for the elderly. Cross-sectional data were collected from 122 wards in health-centre hospitals and residential homes in Finland in 2001. Productive efficiency was measured in terms of technical efficiency, which was defined as the unit's distance from the (best practice) production frontier. The analysis employed stochastic production frontier estimation, where technical inefficiency in the production function was specified to be a function of ward characteristics and clinical quality of care. Several quality indicators based on the Resident Assessment Instrument, such as prevalence of pressure ulcers and depression with no treatment, were used in the analysis. The results did not reveal systematic association between technical efficiency and clinical quality of care. However, the prevalence of pressure ulcers, indicating poor quality of care was associated with technical efficiency, a fact which highlights the importance of including quality measures in the assessment of efficiency in long-term care.

Assessing the efficiency of HIV prevention around the world: methods of the PANCEA project

OBJECTIVE

To develop data collection methods suitable to obtain data to assess the costs, cost-efficiency, and cost-effectiveness of eight types of HIV prevention programs in five countries.

DATA SOURCES/STUDY SETTING

Primary data collection from prevention programs for 2002-2003 and prior years, in Uganda, South Africa, India, Mexico, and Russia.

STUDY DESIGN

This study consisted of a retrospective review of HIV prevention programs covering one to several years of data. Key variables include services delivered (outputs), quality indicators, and costs.

DATA COLLECTION/EXTRACTION METHODS

Data were collected by trained in-country teams during week-long site visits, by reviewing service and financial records and interviewing program managers and clients.

PRINCIPAL FINDINGS

Preliminary data suggest that the unit cost of HIV prevention programs may be both higher and more variable than previous studies suggest.

CONCLUSIONS

A mix of standard data collection methods can be successfully implemented across different HIV prevention program types and countries. These methods can provide comprehensive services and cost data, which may carry valuable information for the allocation of HIV prevention resources.

Frontier efficiency measurement in health care: a review of empirical techniques and selected applications

Health care institutions worldwide are increasingly the subject of analyses aimed at defining, measuring, and improving organizational efficiency. However, despite the importance of efficiency measurement in health care services, it is only relatively recently that the more advanced econometric and mathematical programming frontier techniques have been applied to hospitals, nursing homes, health management organizations, and physician practices, among others. This article provides a synoptic survey of the comparatively few empirical analyses of frontier efficiency measurement in health care services. Both the measurement of efficiency in a range of health care services and the posited determinants of health care efficiency are examined.

Non-parametric and parametric applications measuring efficiency in health care

This paper reviews 188 published papers on frontier efficiency measurement. The techniques used are mainly based on non-parametric data envelopment analysis, but there is increasing use of parametric techniques, such as stochastic frontier analysis. Applications both to hospitals and wider health care areas are reviewed and summarised, and some meta-type analysis undertaken. Results appear to confirm earlier findings that public provision demonstrates less variability than private. The paper is meant as a resource in itself, but also points to the future in terms of possible directions for research in efficiency measurement in health care and health.

How much confidence should we place in efficiency estimates?

Ordinary least squares (OLS) and stochastic frontier (SF) analyses are commonly used to estimate industry-level and firm-specific efficiency. Using cross-sectional data for English public hospitals, a total cost function based on a specification developed by the English Department of Health is estimated. Confidence intervals are calculated around the OLS residuals and around the inefficiency component of the SF residuals. Sensitivity analysis is conducted to assess whether conclusions about relative performance are robust to choices of error distribution, functional form and model specification. It is concluded that estimates of relative hospital efficiency are sensitive to estimation decisions and that little confidence can be placed in the point estimates for individual hospitals. The use of these techniques to set annual performance targets should be avoided.

Health promotion: causes, beliefs and measurements

This investigative study summarizes the five most prevalent healthcare models and seven instruments to help the reader determine which model is the most effective in measuring health-related ideas and behaviors in subjects of varying populations. Their significance to the science and art of health promotion and analytical techniques are also reviewed. The purpose of the study is to consider varying arguments and apply them to abstractions of health promotion activities that readers may be contemplating.

Managed care and technical efficiency

By focusing exclusively on consumer benefit, previous studies of the effects of managed care have ignored important hospital efficiency gains. This study uses the HCUP sample of hospitals for 1992-1996 to estimate a stochastic frontier model of hospital technical efficiency. After controlling for hospital and market area variables, the study finds strong evidence that increased managed care insurance in a given market is associated with improved technical efficiency in the area's hospitals, especially in tertiary cases. Using Battese and Coelli's one-stage method (1995), the coefficients estimates are more efficient than for two-stage methods found in the literature.

Why is there a quality chasm?

Medical care seems to obtain less value from the resources it uses than other industries do, a phenomenon not limited to the United States. I explore several reasons for this, including consumers' ignorance, the rate of technological change, the widespread use of administered pricing, the difficulty of appraising a given provider's quality, and the role of the public sector with objectives other than efficiency. Although these causes suggest that the performance of medical care may always lag behind that of other industries, greater use of information technology and improved financial incentives will help to reduce the size of the quality chasm.

How reliable are hospital efficiency estimates? Exploiting the dual to homothetic production

For scientific use, stochastic frontier estimates of hospital efficiency must be robust to plausible departures from the assumptions made by the investigator. Comparisons of alternative study designs, each well within the 'accepted' range according to current practice, generate similar mean inefficiencies but substantially different hospital rankings. The three alternative study contrasts feature (1) pooling vs partitioned estimates, (2) a cost function dual to a homothetic production process vs the translog, and (3) two conceptually valid but empirically different cost-of-capital measures. The results suggest caution regarding the use of frontier methods to rank individual hospitals, a use that seems to be required for reimbursement incentives, but they are robust when generating comparisons of hospital group mean inefficiencies, such as testing models that compare non-profits and for-profits by economic inefficiency. Demonstrations find little or no efficiency differences between these paired groups: non-profit vs for-profit; teaching vs non-teaching; urban vs rural; high percent of Medicare reliant vs low percent; and chain vs independent hospitals.

Impact of HMO penetration and other environmental factors on hospital X-inefficiency

This study examined the impact of health maintenance organization (HMO) market penetration and other internal and external environmental factors on hospital X-inefficiency in a national sample (N = 1,966) of urban U.S. hospitals in 1997. Stochastic frontier analysis, a frontier regression technique, was used to measure X-inefficiency and estimate parameters of the correlates of X-inefficiency. Log-likelihood restriction tests were used to test a variety of assumptions about the empirical model that guided its selection. Average estimated X-inefficiency in study hospitals was 12.96 percent. Increases in managed care penetration, dependence on Medicare and Medicaid, membership in a multihospital system, and location in areas where competitive pressures and the pool of uncompensated care are greater were associated with less X-inefficiency. Not-for-profit ownership was associated with increased X-inefficiency.

On the efficiency of public, welfare and private hospitals in Germany over time: a sectoral data envelopment analysis study

This paper examines the efficiency of the German hospital sector over time and the relative efficiency of public, welfare (both nonprofit) and private (for-profit) hospital sectors using data from the Federal Statistics Office of German hospitals. Efficiency scores were computed using data envelopment analysis. The absolute efficiency of the hospital sector as a whole was found to have improved between 1991 and 1996. In this comparison, the empirical results showed that the hospitals in the public and welfare sector are relatively more efficient than private hospitals. Our results suggest that public, welfare and private hospital sectors have different best-practice frontiers; and that public and welfare hospital sectors appear to use relatively fewer resources than private hospitals. These results suggest differences in quality of care arising from ownership.

Cost efficiency of US hospitals: a stochastic frontier approach

This study examined the impact of managed care and other environmental factors on hospital inefficiency in 1631 US hospitals during the period 1990-1996. A panel, stochastic frontier regression model was used to estimate inefficiency parameters and inefficiency scores. The results suggest that mean estimated inefficiency decreased by about 28% during the study period. Inefficiency was negatively associated with health maintenance organization (HMO) penetration and industry concentration. It was positively related with Medicare share and for-profit ownership status.

Cost inefficiency in Washington hospitals: a stochastic frontier approach using panel data

We analyze a sample of Washington State hospitals with a stochastic frontier panel data model, specifying the cost function as a generalized Leontief function which, according to a Hausman test, performs better in this case than the translog form. A one-stage FGLS estimation procedure which directly models the inefficiency effects improves the efficiency of our estimates. We find that hospitals with higher casemix indices or more beds are less efficient while for-profit hospitals and those with higher proportion of Medicare patient days are more efficient. Relative to the most efficient hospital, the average hospital is only about 67% efficient.

Alternative methods to examine hospital efficiency: data envelopment analysis and stochastic frontier analysis

There has been increasing interest in the ability of different methods to rank efficient hospitals over their inefficient counterparts. The UK Department of Health has used three cost indices to benchmark NHS hospitals (Trusts). This study uses the same dataset and compares the efficiency rankings from the cost indices with those obtained using Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA). The paper concludes that the methods each have particular strengths and weaknesses and potentially measure different aspects of efficiency. Several specifications should be used to develop ranges of inefficiency to act as signalling devices rather than point estimates. It is argued that differences in efficiency scores across different methods may be due to random "noise" and reflect data deficiencies. The conclusions concur with previous findings that there are not truly large efficiency differences between Trusts and savings from bringing up poorer performers would in fact be quite modest.

Efficiency and administrative costs in primary care

We use a formal model to examine the implications of endogenous managerial effort for the interpretation and estimation of efficiency in health care organisations. The model is applied to investigate the doubling of the cost of administering primary care in England in real terms between 1989/1990 and 1994/1995. The main cost determinant was the number of general practitioners (GPs), and there were economies of scale but not of scope. Fund-holding had a positive but small effect on administrative costs, so that the recent abolition of fund-holding may do little to reduce primary care administrative costs. After allowing for changes in the numbers of primary care practitioners, the quality of primary care and the extent of fund-holding, most of the increase in costs was unexplained, and may reflect additional but unmeasured increases in the administrative burden associated with the 1990 reforms. There was little variation in relative efficiency across areas.

A multilevel modelling approach to analysis of patient costs under managed care

The growth of the managed care model of health care delivery in the USA has led to broadened interest in the performance of health care providers. This paper uses multilevel modelling to analyse the effects of managed care penetration on patient level costs for a sample of 24 medical centres operated by the Veterans Health Administration (VHA). The appropriateness of a two level approach to this problem over ordinary least squares (OLS) is demonstrated. Results indicate a modicum of difference in institutions' performance after controlling for patient effects. Facilities more heavily penetrated by the managed care model may be more effective at controlling costs of their sicker patients.

Graduate medical education costs in nonacademic health center teaching hospitals: evidence from Maryland

As managed care has grown, much concern has been expressed about the potential plight of the nation's 125 academic health centers (AHCs). Less concern has focused on non-AHC teaching hospitals, although most studies of graduate medical education (GME) costs include these hospitals in their estimates. While most studies have found that costs increase positively with various measures of "teaching intensity," some have concluded that hospitals with smaller programs have costs that are the same or less than comparable nonteaching hospitals. However, few studies have tested whether AHCs' cost structures are sufficiently similar to those of other hospitals to reliably include them in the same estimation. This article tests that assumption for Maryland hospitals, finds it violated, and presents results for non-AHC teaching hospitals. The results reveal that, at least in Maryland, even small teaching programs add to hospital costs.

Data envelopment analysis to determine efficiencies of health maintenance organizations

We use Data Envelopment Analysis (DEA) to measure the relative technical efficiencies of 164 HMOs licensed to practice in the United States in 1995 with data collected from the American Association of Health Plans. Health care output measures used in the analysis are the number of commercial, Medicare and Medicaid lives covered in each plan. Inputs to the model are health care utilization measures such as the number of medical and surgical inpatient days, number of maternity and newborn stays in days, number of outpatient and emergency room visits and the number of non-invasive and invasive procedures performed on patients in an ambulatory setting. Mean efficiency of health plans was 40% (of the most efficient). We use multivariate analysis to try and explain variations in efficiency. Enrollment influences efficiency, with larger HMOs being more efficient than those with fewer enrollees. Plans with a more even distribution of Commercial, Medicare and Medicaid patients were more efficient on average than plans with heterogeneous mixes in enrollment. HMOs with Medicare patients are significantly less efficient, with efficiency decreasing with increasing Medicare participation in plan membership. Health plans in operation for longer periods of time had greater outputs with the same inputs. Health plans that had a majority of their enrollees in network or IPA type arrangements were more efficient as were for-profit plans compared to not-for-profits. Policy implications are discussed.