Chapter 6. Efficiency

Few issues within American health care policy today are as extensively debated as how to obtain better value for money. The debate about how to improve efficiency is equally matched by the debate about how best to measure it. Varying perspectives and definitions of "efficiency" in the health care marketplace and the lack of consensus on what constitutes appropriate measurement of efficiency have stymied efforts to report on this area. For example, efficiency can be viewed from different perspectives, including individual patients, providers, and society as a whole.

The issue of how to improve efficiency in the Nation's health care system is at the heart of the Department of Health and Human Services' (HHS) mission to increase transparency in health care through better information on quality and cost. In support of this mission, this year's National Healthcare Quality Report (NHQR) continues to look at potential information sources and findings on efficiency in the U.S. health care system.

This year's NHQR outlines the varying perspectives of efficiency and offers potential methods for measuring efficiency at the national level that respond to the NHQR's mandate to provide lawmakers in Congress with information on health care performance. This chapter does not attempt to provide a definitive framework for efficiency, nor does it provide an exhaustive list of potential measures of efficiency. The examples follow the initial effort to report on efficiency in the 2007 NHQR and should still be viewed as preliminary.

No conclusions about efficiency in the U.S. health care system should be drawn. Rather, the Agency for Healthcare Research and Quality (AHRQ) hopes that this chapter will stimulate further productive discussions on health care efficiency. AHRQ intends this chapter to be part of an evolving national discussion on measuring efficiency in the U.S. health care system that will be reviewed, revised, and presented in future reports.

Return to Contents

Background and Measures

In its landmark report, Crossing the Quality Chasm: A New Health System for the 21st Century,¹ the Institute of Medicine (IOM) presented six "aims" for the health care system: effectiveness, safety, timeliness, patient centeredness, equity, and efficiency. AHRQ, in its 2001 reauthorization legislation, was given the task of developing two national health care reports that would track quality and prevailing disparities in the Nation's health care system.

IOM provided guidance^2,3 on the development of these two national health care reports and suggested that the reports' framework be linked to the six aims presented in Crossing the Quality Chasm. At the same time, however, IOM stated that AHRQ should not try to address the issue of efficiency in the first national reports but should examine its inclusion in future reports or in a separate report.

With guidance from an HHS Interagency Work Group brought together to advise on the reports' development, AHRQ developed the first NHQR and National Healthcare Disparities Report (NHDR) in 2003 without addressing efficiency. In 2004, the Interagency Work Group encouraged AHRQ to examine possible approaches to including efficiency in future reports. This followed advice from AHRQ's National Advisory Committee (NAC) of external experts from the private sector, academia, and the Federal sector. The NAC had, at AHRQ's request, formed a subcommittee, led by Dr. Don Berwick, that provided advice on the NHQR and NHDR. That subcommittee recommended that AHRQ develop a chapter on efficiency for the reports.

To respond to the NAC and Interagency Work Group requests, AHRQ formed a subgroup of its Interagency Work Group in 2004 to address efficiency. In 2005, this subgroup held two meetings, during which it reviewed documents from previous reports and discussed possible ways to further this effort. The subgroup concluded that there was insufficient consensus to conceptualize and measure efficiency.

AHRQ had previously commissioned the RAND Corporation to systematically review measures of efficiency and their potential to be tracked and reported at various levels. The efficiency subgroup therefore decided to wait until RAND submitted its report to AHRQ before developing any further plans. The final version of the RAND report summarizes the knowledge base on efficiency measures as follows:

• Few analyses of the reliability and validity ("scientific soundness") of published and unpublished measures have been conducted.
• Both the published literature measures and the vendor measures focus on intermediate outcomes (e.g., inpatient stays), not final outcomes (e.g., functional status or measures of health).
• Consensus has yet to emerge on which approaches constitute acceptable measures of efficiency.⁴

The RAND report provides a typology of efficiency measures that emphasizes the multiple perspectives on efficiency and points out that measures must be considered from the standpoint of what the measuring organization is and what its goal is in assessing efficiency. The typology distinguishes between:

• Society as a whole (i.e., the "population" level).
• Health care firms (i.e., hospitals and other providers).
• Individuals.

Another report in 2006 examined the question of efficiency from the cost of waste point of view. In that report, the authors outline another common typology for efficiency measurement: the tracking of overuse, underuse, and misuse in the health care system.⁵

This chapter first presents a general set of trends on costs and quality levels in the U.S. health care system. Additional measures summarize information at the population and provider level. The measures used are presented to provide some insight on health care efficiency. They are:

• Change in expenditures and quality of care for cancer, diabetes, and heart disease (overview).
• Trends in avoidable hospitalizations and costs (population perspective).
• Rehospitalization for congestive heart failure (CHF) for selected States (population perspective).
• Trends in hospital efficiency (provider perspective).

Because consensus has yet to emerge about the appropriate framework and acceptable measures of efficiency, the examples provided should be viewed as preliminary and designed to stimulate productive ongoing discussions about health care efficiency.

Return to Contents

Findings

Change in Expenditures and Quality of Care for Cancer, Diabetes, and Heart Disease

Data from AHRQ's Medical Expenditure Panel Survey (MEPS) are used to provide a preliminary overview and to suggest possible national trends in health care cost and quality. MEPS collects health care expenditures by all payers for nearly all types of health care utilization, including outpatient visits, hospital inpatient stays, emergency department visits, prescribed medicines, dental visits, and home health care. Data are collected for the civilian noninstitutionalized population.

Summary data are presented here on the average annual rate of change from 2001 to 2005 in total annual expenditures for the general population and for people with three high-priority conditions: cancer, diabetes, and heart disease. In addition, quality data are summarized in terms of the median rate of change in the NHQR measures from 2001 to 2005 for the entire measure set and for each condition area.ⁱ

Figure 6.1. Average annualized percentage changes in national health care expenditures and quality for general population and people with selected conditions,^* 2001-2005

^* Refer to Chapter 1, Introduction and Methods, for a discussion of how data years were selected for determining the percentage change in health care quality.
Source: Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey (MEPS), 2001 and 2005. Refer to the Measure Specifications appendix for list of measures included in each category.
Reference population: Civilian noninstitutionalized population.
Note: Expenditures are payments from all sources for hospital inpatient care, ambulatory care provided in offices and hospital outpatient departments, care provided in emergency departments, and prescribed medicine purchases reported by respondents in the MEPS-Household Component. Sources include direct payments from individuals, private insurance, Medicare, Medicaid, Workers' Compensation, and other miscellaneous sources. Expenditures for 2001 are adjusted to 2005 dollars using the gross domestic product implicit price deflator (Bureau of Economic Analysis).

• From 2001 to 2005, total annual health care expenditures increased at a rate 4.6 times the rate of the increase in the summary measure of quality of care. Annual total health care expenditures rose 6.5% (in 2005 dollars). During this same period, quality increased at a rate of 1.4% (Figure 6.1).
• For heart disease, cancer, and diabetes individually, quality increased at a rate of 2.6%, 1.9%, and 0.1% annually, respectively. Expenditures increased at an annual rate of 4.4%, 9.0%, and 4.9%, respectively.

Figure 6.1 may seem to suggest that improvements in overall quality are outpaced by increases in expenditures. However, such a conclusion cannot be drawn, and the statistics should be viewed with caution because these are comparisons of percentage changes in two very different measures. First, expenditures are comprehensively measured, but quality is not. Figure 6.1 presents a summary of all available quality measures in this report rather than a catalog of all clinical care for all conditions and patients. The quality measures track both processes of care and outcomes of care.

The indicators selected for inclusion in the NHQR and NHDR measure set are considered the most scientifically sound and clinically important markers of whether we are achieving appropriate performance in health care. However, many aspects of care are not captured in these quality indicators. A comprehensive assessment may never be feasible, as technical aspects of care are changing more rapidly than can be captured through broad, consensus-based quality measurement vehicles, such as the NHQR.

Moreover, it would be difficult to collect measures of quality for rare conditions. In addition, the summary measure of quality is composed of measures calculated on a per person basis, but total annual expenditures increase in part due to population growth. Finally, these statistics are provided without estimates of variability (i.e., without confidence intervals). Statistical testing for these sorts of comparisons is complex, and future versions of the NHQR will examine more refinements to such statistical testing.ⁱⁱ

The statistics illustrated above suggest many questions about efficiency. They are not provided to suggest causation between costs and quality. Providing higher quality care may cost more than providing lower quality care, and achieving increasingly higher quality goals may require even higher expenditures to reach an additional person. Some types of quality care might reduce expenditures, particularly by reducing hospitalizations. Furthermore, the factors that cause changes in expenditures may be different from the factors that cause improvements in quality. More research is needed to investigate these issues.

Trends in Avoidable Hospitalizations and Costs

To address the population perspective of potentially avoidable hospitalizations and costs, data on ambulatory-care-sensitive conditions are summarized here using the AHRQ Prevention Quality Indicators (PQIs). Not all hospitalizations that the AHRQ PQIs track are preventable, but ambulatory-care-sensitive conditions are those for which good outpatient care can prevent the need for hospitalization or for which early intervention can prevent complications or more severe disease.⁶ The AHRQ PQIs track these conditions using hospital discharge data. For this analysis, total hospital charges were converted to costs using Healthcare Cost and Utilization Project (HCUP) cost-to-charge ratios based on hospital accounting reports from the Centers for Medicare & Medicaid Services. Therefore, cost estimates in this section refer to hospital costs.

Figure 6.2. National trends in potentially avoidable hospitalization rates, by type of hospitalization, 1997 and 2000-2005

Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project, Nationwide Inpatient Sample, 1997 and 2000-2005.
Note: Data are for adults age 18 and over. Annual rates are adjusted for age and gender.

 

Figure 6.3. Total national costs associated with potentially avoidable hospitalizations, 1997 and 2000-2005

Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project, Nationwide Inpatient Sample, 1997 and 2000-2005.
Note: Data are for adults age 18 and over.

• From 1997 to 2005, avoidable hospitalizations for chronic conditions decreased significantly, from 1,294 per 100,000 to 1,092 per 100,000 (Figure 6.2). The rate also decreased significantly from 2000 to 2005 (1,213 per 100,000 to 1,092 per 100,000).
• Avoidable hospitalizations for acute conditions did not significantly change from 1997 to 2005 or from 2000 to 2005.
• Although avoidable hospitalization rates have decreased overall since 2000, total national costs associated with potentially avoidable hospitalizations have increased since 2000 (Figure 6.3). Hospital costs due to avoidable hospitalizations exceeded $29 billion in 2005, which was 35% greater than what these costs were in 1997 when adjusted for inflation ($21.9 billion).ⁱⁱⁱ

These figures provide some preliminary measures of the potential for improvement in one dimension of efficiency.

Rehospitalization for Congestive Heart Failure

To gain further insight into the population perspective of avoiding potentially avoidable hospitalizations and costs, data on rehospitalization rates for CHF for nine States in 2004 and 2005 are summarized here (Table 6.1). Rehospitalization for CHF signals a worsened state of illness for patients and is more resource intensive than treatment as an outpatient in the community. Although not every rehospitalization for CHF is preventable, CHF is a condition for which good outpatient care and early intervention can help prevent rehospitalization.

The estimates below are derived from data for nine States participating in the HCUP State Inpatient Databases. They are based on all CHF admissions from January to September of each year and allow for a 3-month timeframe for rehospitalization.^iv Rehospitalizations have a principal diagnosis of CHF.

Table 6.1. Rehospitalizations for congestive heart failure, per 1,000 initial admissions for CHF, 9 States, 2004 and 2005

Age category	State	2004		2005
		Rate	SE	Rate	SE
Ages 18-64	State A	140	1.00	140	1.00
	State B	190	1.00	230	1.00
	State C	180	1.00	190	1.00
	State D	260	0.00	270	0.00
	State E	230	1.00	220	1.00
	State F	220	1.00	240	1.00
	State G	230	1.00	230	1.00
	State H	240	0.00	250	0.00
	State I	240	1.00	250	1.00
Ages 65+	State A	110	1.00	110	1.00
	State B	170	0.00	160	0.00
	State C	180	0.00	170	0.00
	State D	190	0.00	190	0.00
	State E	210	0.00	200	0.00
	State F	210	0.00	200	0.00
	State G	200	1.00	200	0.00
	State H	210	0.00	210	0.00
	State I	220	0.00	210	0.00
All ages (18+)	State A	120	1.00	120	1.00
	State B	170	0.00	180	0.00
	State C	180	0.00	180	0.00
	State D	210	0.00	210	0.00
	State E	210	0.00	200	0.00
	State F	210	0.00	210	0.00
	State G	210	0.00	210	0.00
	State H	220	0.00	220	0.00
	State I	220	0.00	220	0.00

Key: SE = standard error.
Source: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project, State Inpatient Databases, 2004 and 2005.
Note: Data are for adults age 18 and over.

• The mean CHF rehospitalization rate for all adult patients previously admitted for CHF in the 9-State sample was 210 per 1,000 in both 2004 and 2005.
• The rate of State-level CHF rehospitalizations for adult patients of all ages previously admitted for CHF ranged from a low of 120 to a high of 220 per 1,000 for rehospitalizations for CHF. State A had a rehospitalization rate that was consistently lower than the other rates across both years and all reported age categories.
• The CHF rehospitalization rates for patients previously admitted for CHF were generally lower in the Medicare-eligible population than in those ages 18-64.

It is important to note that the figures reported above are not national estimates and that no conclusions about national trends should be inferred. The States in the analysis account for about 36% of all adult discharges for CHF and provide an indication of the general trend that readmissions for CHF may be following.

Trends in Hospital Efficiency

Significant attention has been paid to cost variations across providers and across the country. Yet it is often difficult to separate out costs due to differences among providers in outputs, patient burden of illness,⁵ or care quality. To address the provider perspective, hospital cost efficiency is examined using a technique from the field of econometrics that can account for such differences.^v This analysis uses data from the American Hospital Association Annual Survey, Medicare Cost Reports, and HCUP State Inpatient Databases.

Here, hospital efficiency is defined as the ratio of best practice costs to total observed costs. For example, given the types and quantities of outputs a hospital produces, the input prices it pays, its case mix, its quality, and its market characteristics, a theoretical best practice hospital might incur expenses amounting to $90 million. A comparison hospital in an identical situation with total expenses of $100 million would have an estimated cost efficiency of 90%.

Cost-efficiency estimates have been converted to index numbers with a base of 100 for the year 2001 as a way to place less emphasis on the specific magnitude of estimated cost efficiency than on its general trend.

This analysis controls for the following components that Elixhauser, et al. (1998) contend are part of patient burden of illness: (1) primary reason for admission to the hospital, (2) severity of the principal diagnosis, (3) iatrogenic complications, and (4) comorbidities that are unrelated to the primary diagnosis but have a substantial impact on both the resources used to treat the patient and the outcomes of the care provided.⁸

Figure 6.4. Average estimated relative hospital cost efficiency index for a selected sample of urban general community hospitals, 2001-2005

Source: Agency for Healthcare Research and Quality. Analysis based on 1,368 urban general community hospitals with data in the Healthcare Cost and Utilization Project, State Inpatient Databases. Go to Chapter 1, Introduction and Methods, for further details.

• Estimated urban hospital cost efficiency increased slightly from 2001 to 2004 but decreased slightly in 2005 for a selected sample of urban general community hospitals (Figure 6.4)
• The most cost-efficient hospitals (i.e., hospitals in the highest quartile of estimated cost efficiency) compared favorably with the least cost-efficient hospitals (i.e., hospitals in the lowest quartile of estimated cost efficiency) on a number of important variables. The most cost-efficient hospitals had lower costs and fewer full-time-equivalent employees per case-mix-adjusted admission, as well as a shorter average length of stay, compared with the least cost-efficient hospitals (Table 6.2).
• The most cost-efficient hospitals had a higher operating margin than the least cost-efficient hospitals (Table 6.2).

Table 6.2. Correlates of hospital cost efficiency

Measure	Estimate	Standard deviation
Cost per case-mix-adjusted admission:
Top quartile of hospital cost efficiency	$4,340	$1,087
Bottom quartile of hospital cost efficiency	$6,241	$2,350
Full-time equivalent employees per case-mix-adjusted admission:
Top quartile of hospital cost efficiency	.040	0.01
Bottom quartile of hospital cost efficiency	.055	0.02
Average length of stay (days):
Top quartile of hospital cost efficiency	4.88	1.33
Bottom quartile of hospital cost efficiency	5.22	1.80
Operating margin:
Top quartile of hospital cost efficiency	.033	0.13
Bottom quartile of hospital cost efficiency	-.066	0.17

Source: American Hospital Association Annual Survey of Hospitals and Medicare Cost Reports, 2001-2005.

It is important to note that the figures reported above are not national estimates and that no conclusions about national trends should be inferred. However, the hospitals in the analysis represent about 53% of all urban general community hospitals and therefore provide an indication of the general trend that cost efficiency may be following.

---

ⁱ This median rate of change is the same metric used in the Highlights section of this report and is explained in detail in Chapter 1, Introduction and Methods. A list of the measures used for these calculations is available in the Measure Specifications appendix.
ⁱⁱ The creation of confidence intervals for expenditures using MEPS data is possible and was conducted for this analysis. The estimates with their confidence intervals are: (1) heart disease, 4.4% (-2.1-10.9); (2) cancer 9.0%, (0.3-17.6); and (3) diabetes, 4.9% (-1.4-11.2).
ⁱⁱⁱ The inflation adjustment was done using the gross domestic product implicit price deflator.
^iv A 3-month period (or longer) for readmissions has been used in studies of the effective management of chronic illness.⁷ Two-thirds of the readmissions reported above occurred within 1 month.
^v Stochastic frontier analysis (SFA) is the technique used in this analysis. SFA can estimate best practice costs as the value total costs would be if full efficiency were attained. The hospital-level "cost efficiency" estimates SFA produces measure whether output is obtained using the fewest inputs (i.e., technical efficiency), as well as whether output is produced using the optimal mix of inputs, given prices (i.e., allocative efficiency), the size of a hospital's operations (i.e., scale efficiency), and the range of a hospital's operations (i.e., scope efficiency), including possible overspecialization or overdiversification.⁹

---

Return to Contents

Next Steps in Efficiency Reporting

A significant amount of information about about the study of health care efficiency and its measurement is not fully developed. In addition, the relationship between health care quality and efficiency is complex and not well understood. Recent work examining variations in Medicare spending and quality shows that higher cost providers do not necessarily provide higher quality care, illustrating the potential for improvement.¹⁰ The preliminary examination of efficiency in this chapter is only an early step. Tracking efficiency in the health care system over the long term, understanding its relationship with quality, and finding ways to improve quality and efficiency will require future research commitment in these areas.

The AHRQ-sponsored and commissioned report prepared by RAND, Identifying, Categorizing, and Measuring Health Care Efficiency Measures, released in April 2008, identifies several gaps in efficiency measurement.⁴ A number of efforts are now underway to advance our knowledge of efficiency. On May 20, 2008, AHRQ hosted the "Physician Performance Measurement and Reporting Conference," cosponsored by the Wisconsin Collaborative for Healthcare Quality. This conference included more than 40 technical experts on physician performance measurement and stakeholders representing consumers, purchasers, employers, and providers. The goal of the meeting was to identify areas of agreement across the groups and to suggest topics where future research might lead to further consensus. Topics identified as promising areas for future research include:

• Can standards be developed that can be applied to the diverse physician practices seen across the United States?
• Can a bridge between claims and clinical data be created?
• Will the implementation of initiatives designed to improve health care efficiency have an impact on health care quality?
• How can the regional variations in coding practices be addressed?
• How does the size of a physician practice affect the measurement of performance?

The RAND report notes that efficiency measurement techniques developed in the academic literature, namely "frontier techniques," have not been applied to the policy setting. The RAND report identified these techniques, including stochastic frontier analysis, which is used to provide the population perspective in this chapter, as among the most promising approaches for measuring provider efficiency.

These approaches also may inform strategies for improving the delivery of health care services. Therefore, AHRQ hosted an invitational meeting, "Translating Frontiers Into Practice: Taking the Next Steps Toward Improving Efficiency," on August 27-28, 2008. This meeting brought together policymakers, stakeholders, and leading technical experts to discuss how frontier techniques can be used most effectively to address the problems confronting the health care system and to identify how the needs of end users should shape the research community's agenda.

Moreover, AHRQ sponsored the first in what will be a series of theme issues of the journal Health Services Research. This first issue, which appeared in October 2008, is called "Improving Efficiency and Value in Health Care." It provides insight into recent initiatives in health care that require efficiency measurement. These efforts range from internal quality improvement exercises to innovations in payment and public reporting. The theme issue also emphasizes the importance of organizational structures and market forces to efforts aimed at improving efficiency and increasing value.

One of the primary areas on which AHRQ and its HHS partners are concentrating on improving efficiency measurement is the Secretary's Value-Driven Health Care Initiative. This initiative is an effort by Secretary Leavitt and HHS to provide public information about the quality and cost of services health care providers deliver. Such information is not widely available today; thus, there is little information to help consumers compare doctors and hospitals based on measures of quality and cost. Providers themselves have limited information for comparing their performance based on accepted standards of care. Yet such information may be crucial for delivering the best treatment and the best value in health care.

As part of the Value-Driven Health Care Initiative, volunteer participants in AHRQ Chartered Value Exchanges (http://www.hhs.gov/valuedriven/communities/exchanges.html) commit to four objectives, called the "cornerstones" of value-driven health care. One cornerstone is "Measure and Publish Quality Information," whereby participants commit to public reporting on the performance of doctors, hospitals, and other providers. The other cornerstones are "Interoperable Health Information Technology," "Measure and Publish Price Information," and "Promote Quality and Efficiency of Care."

For more information about the Value-Driven Health Care Initiative, visit http://www.hhs.gov/valuedriven/.

Return to Contents

References

1. Institute of Medicine. Committee on Quality of Health Care in America. Crossing the quality chasm: a new health system for the 21st century. Washington, DC: National Academy Press; 2001.

2. Institute of Medicine. Envisioning the National Healthcare Quality Report. Washington, DC: National Academy Press; 2001.

3. Swift EK, ed. Institute of Medicine. Committee on Guidance for Designing a National Healthcare Disparities Report. Guidance for the National Healthcare Disparities Report. Washington, DC: National Academies Press; 2002.

4. McGlynn E. Identifying, categorizing, and evaluating health care efficiency measures. Final Report. (Prepared by Southern California Evidence-based Practice Center" RAND Corporation under Contract No. 282-00-0005-21). Rockville, MD: Agency for Healthcare Research and Quality; April 2008. AHRQ Publication No. 08-0030. Available at: http://www.ahrq.gov/qual/efficiency/.

5. Cost of poor quality or waste in integrated delivery system settings. Final Report. (Prepared by RTI International under Contract No. 290-00-0018-11). Rockville, MD: Agency for Healthcare Research and Quality; September 2006. Available at: http://www.ahrq.gov/research/costpoorids.pdf.

6. AHRQ Quality Indicators. Guide to Prevention Quality Indicators: hospital admission for ambulatory care sensitive conditions. Rockville, MD: Agency for Healthcare Research and Quality; March 12, 2007. Version 3.1. Available at: http://www.qualityindicators.ahrq.gov/pqi_download.htm.

7. Fries K, Koop E, Sokolov J, et al. Beyond health promotion: reducing need and demand for medical care. Health Aff 1998;17(2):70-84.

8. Elixhauser A, Steiner C, Harris R, et al. Comorbidity measures for use with administrative data. Med Care 1998;36:8-27.

9. Mutter R, Rosko M, Wong H. Measuring hospital inefficiency: the effects of controlling for quality and patient burden of illness. Health Serv Res 2008;43:1992-2013.

10. Fisher E, Wennberg D, Stukel T, et al. The implications of regional variations in Medicare spending. Part 1: the content, quality, and accessibility of care. Ann Intern Med 2003;138:273-87.

Return to Contents