A.5 Prototype Indicator Set: Health Care Resource Allocation

BACKGROUND

Since the 1950s, the share of gross national product devoted to personal health care has increased from 6 percent to 14 percent in the United States. This major annual investment has achieved substantial gains in the treatment of many diseases and has almost certainly contributed to improvements in life expectancy (Bunker et al., 1995). However, the United States still lags many other countries in critical measures of population health and of health system performance. For example, about 40 million Americans lack health insurance (Summer, 1994). Given that the United States spends almost 50 percent more per capita on health care than many other developed countries, the principal barrier cannot be a lack of resources to meet these needs. Rather, the United States faces a challenge of how to allocate its resources more efficiently to improve the health of the entire population.

Efficiency in health care entails achieving greater value—improved health status and increased satisfaction for a given expenditure. Although much work remains to be done to define value in health care, evidence of several types of inefficiency can be found in our current system.

First, there are numerous examples of the provision of ineffective or unwanted care (i.e., waste). The Prostate Patient Outcome Research Team, for example, found that many surgical interven-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring A.5 Prototype Indicator Set: Health Care Resource Allocation BACKGROUND Since the 1950s, the share of gross national product devoted to personal health care has increased from 6 percent to 14 percent in the United States. This major annual investment has achieved substantial gains in the treatment of many diseases and has almost certainly contributed to improvements in life expectancy (Bunker et al., 1995). However, the United States still lags many other countries in critical measures of population health and of health system performance. For example, about 40 million Americans lack health insurance (Summer, 1994). Given that the United States spends almost 50 percent more per capita on health care than many other developed countries, the principal barrier cannot be a lack of resources to meet these needs. Rather, the United States faces a challenge of how to allocate its resources more efficiently to improve the health of the entire population. Efficiency in health care entails achieving greater value—improved health status and increased satisfaction for a given expenditure. Although much work remains to be done to define value in health care, evidence of several types of inefficiency can be found in our current system. First, there are numerous examples of the provision of ineffective or unwanted care (i.e., waste). The Prostate Patient Outcome Research Team, for example, found that many surgical interven-

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring tions were being recommended by physicians based on symptom level alone, rather than on the degree to which the symptoms actually bothered patients or an understanding of the risks and benefits of surgery (Wasson et al., 1993). Providing balanced information led to substantial reductions in the rate of surgery. Investigators from RAND found that many of the diagnostic tests and surgical procedures that are performed are inappropriate (e.g., Park et al., 1989). The use of hospitals when alternative, less costly, and less intrusive sites of care are equally effective represents a similar problem—the unnecessary consumption of resources. A second form of inefficiency can be found in failures to provide care that is known to be effective (i.e., lost opportunities). The failure to provide low-cost interventions of well-documented efficacy and effectiveness would represent an inefficiency of the community health system. From this perspective, the failure to provide childhood immunizations, adequate prenatal care, or appropriate screening for the early diagnosis of cervical cancer would all be evidence of an inefficient allocation of health care resources. A third form of inefficiency is found when one examines differences in expenditures for the performance of specific health care-related tasks in a defined population (i.e., process inefficiency). Given an equal outcome, the treatment provided at a lower cost to the consumer is more efficiently delivered. This raises questions about both the amount of resources used and the prices paid for those services. Evidence of differences in the prices paid to providers for health care services is substantial for hospital services (e.g., Pennsylvania Health Care Cost Containment Council, 1995), for physician services (Welch et al., 1996), and for the provision of specific high-cost procedures that entail both physician and hospital care. Such evidence has led many purchasers to contract with specific ''centers of excellence" that can provide high-quality care at a lower price—an improvement in efficiency from the payer's perspective. Consideration of price also leads quickly into highly charged questions about what represents a reasonable profit or operating margin. What is a fair return for a hospital? What is a reasonable income for a physician? What is a fair profit for a health plan? Finally, one can consider outcome inefficiency, or expending more resources than necessary to achieve a given health outcome in a defined population. An approach to defining populations based in the current market-driven reforms of the health care system is the effort led by the National Committee for Quality

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Assurance to evaluate enrollees' experience of managed care plans. HEDIS, the Health Plan Employer Data and Information Set (NCQA, 1993), encompasses measures in several categories, including quality of care for both prevention services and treatment, utilization of services, members' access to and satisfaction with services, and organization and operation of the health plan. Measures, chosen to produce information that promotes quality improvement within health plans, are reviewed and revised on a periodic basis. A second population-based approach is reflected in the geographic analyses that have documented large variations in per capita health care resources, utilization, and expenditures across U.S. communities (e.g., Wennberg, 1996). Although there are clearly differences in population characteristics across geographic areas, the differences observed in the use of health care services have not been explained by detectable differences in the need for services (Wennberg, 1996), nor by any evidence that greater resource use is associated with improved health outcomes or satisfaction with care. A comparison of population-based outcomes of acute myocardial infarction (AMI) demonstrated that lower rates of post-AMI intervention were associated with both improved survival and improved control of symptoms (Guadagnoli et al., 1995). Another demonstration of inefficient resource allocation is evident in a comparison of Medicare beneficiaries in Miami and Minneapolis. Although health care spending for Medicare beneficiaries residing in Miami is twice that for residents of Minneapolis (i.e., $5,922 per beneficiary compared to $2,966, after adjustment for differences in age, sex, race, and price), mortality rates are identical (i.e., 47 per 1,000 beneficiaries). A recent survey also found that residents of Miami are less satisfied with their care (Wennberg, 1996; J. Knickman, personal communication, 1996). "FIELD" SET OF PERFORMANCE INDICATORS The field model encourages a shift of focus from individual patients or enrolled populations to the community as a whole. Moreover, the model explicitly raises the question of whether the marginal investment in personal health care would not achieve a greater benefit if invested in other sectors of the community. The potential stakeholders for such an effort include all segments of the community. A community-based focus also draws attention to two problems that are easily ignored when the focus remains at the level either of the individual patient or of enrolled populations.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring The first is the problem of the uninsured or other disadvantaged populations who may have difficulty receiving basic services. The second is the issue of the responsibility of local health care institutions to the community itself (Kaufman and Waterman, 1993; Showstack et al., 1996). The committee used the field model as a framework to guide the selection of potential indicators of inefficient allocation of health care resources. The committee also recognized the importance of cross-community or longitudinal comparisons as guides to the interpretation of resource indicators. For many of the indicators, the absolute level of resources required to meet population health needs is unknown. Benchmarking to other communities that are similar in terms of measures of need and yet appear to meet those needs with fewer resources provides a plausible and rational justification for resource reallocation (Fisher et al., 1992). The selection of performance indicators of health care resource allocation was thus guided both by the field model and by the possibility of applying an approach based on benchmarking. Four domains of the field model are particularly useful for developing measures of potential need for health care within a community: disease, health and function, prosperity, and well-being. The committee suggests that data from its proposed community profile be assembled on the target community and on potential benchmark communities. Within the remaining domains, potential indicators are identified below. Individual Behavior The field model highlights the influence of factors in the environment on individual behavior. Examining the proportion of inpatient deaths that occur in the absence of a completed advance directive acknowledges that hospitals and physicians have an important role in influencing this specific individual behavior. Patients with advance directives are significantly more likely to participate in end-of-life decisions and to limit medical treatment when facing a terminal disease (Weeks et al., 1994). The committee proposed this measure of potential inefficiency because of the potentially high costs that may be incurred in terminal illness in which care continues only because an advance directive was not completed (Emanuel, 1996). A second measure of the impact of factors in the environment that influence individual behavior is the rate of cigarette smoking.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Smoking is a significant underlying risk factor for many adverse health conditions (McGinnis and Foege, 1993). Physicians and many other individuals and organizations in the community can influence individuals' smoking behavior. Social Environment The excess capacity of the U.S. hospital system is well recognized, but the challenges of downsizing are substantial. A barrier to reallocation of resources from acute care to other sectors is the widespread assumption that greater levels of investment in health care are beneficial. The evidence of benefit from greater spending on personal health care is limited (Bunker et al., 1995), and there is some evidence to the contrary (Guadagnoli et al., 1995). The prevalence of this assumption is difficult to measure directly. However, the media portrayal of health care and public health services represents a potential construct for examining social assumptions regarding the value of health care spending. The data are not currently available. Physical Environment The physical environment of health care may provide some evidence of inefficiency. Two examples considered by the committee focus on regionalization. Substantial evidence documents the improved outcomes and lower costs that are achieved when patients receive coronary artery bypass graft surgery in high-volume regional centers (Luft et al., 1990; Hannan et al., 1991). Similarly, because most high-risk deliveries should be identified before birth, inadequate regionalization of high-risk obstetrical care can be identified by the number of low birth weight infants born in centers without appropriate facilities to handle them. Communities may want to examine the number of avoidable hospitalizations by measuring ambulatory care sensitive (ACS) hospitalizations, an indirect measure of underutilization of outpatient services and poor primary care management (IOM, 1993). Studies correlating ACS hospitalizations with other factors are complex, but a raw count of inpatient hospitalizations for conditions that are known to be manageable in an outpatient setting (e.g., diabetes, asthma) might be adequate for communities. Data may be available through HEDIS and through state comprehensive hospital discharge databases (e.g., the Washington State Commission Hospital Abstract Reporting System), where available.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Health Care Within the domain of health care, the committee identified three constructs of particular importance in judging the efficiency of resource allocation: (1) underprovision of basic, cost-effective services (i.e., missed opportunities); (2) overprovision of unnecessary services (i.e., waste); and (3) excess capacity (i.e., outcome inefficiency). The specific measures proposed for each of these constructs can feasibly be obtained from sources listed in the accompanying table. Inadequate prenatal care. A reasonable indicator of inadequate prenatal care is the percentage of births to women who did not receive any care during the first trimester. The data for this measure are readily available through state vital statistics. Inadequate primary care. The percentage of the population without a regular source of care, other than emergency departments, can be ascertained through the Behavioral Risk Factor Surveillance System (BRFSS) or may be available through other surveys as well. Communities may consider measuring selected relevant clinical outcomes. For example, COMAH (Clinical Outcome Measure Adjusted HEDIS) indicators are being pilot tested in the state of Washington (J. Krieger, personal communication, 1996). These indicators attempt to measure "root issues" such as respiratory flow for asthmatics, immunization levels for children, and blood sugar levels for diabetics. Inadequate public health capacity. Public health is an essential component of the national health system. Healthy People 2000 (USDHHS, 1991) includes an objective calling for 90 percent of the population to be served by a local health department that effectively caries out the core public health functions of assessment, policy development, and assurance. Essential services to fulfill those core functions have been defined as (Baker et al., 1994): Monitor health status to identify and solve community health problems. Diagnose and investigate health problems and health hazards in the community. Inform, educate, and empower people about health issues.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Mobilize community partnerships and action to identify and solve health problems. Develop policies and plans that support individual and community health efforts. Enforce laws and regulations that protect health and assure safety. Link people to needed personal health services and assure provision of health care when otherwise unavailable. Assure a competent workforce for public health and personal care. Evaluate effectiveness, accessibility, and quality of personal and population-based health services. Research for new insights and innovative solutions to health problems. Efforts are being made to develop tools that communities can use to evaluate the adequacy of the public health system in their local areas. A set of 10 public health practices have been used as a basis for developing survey instruments for assessing local health department effectiveness in performing the core functions (Miller et al., 1994; Turnock et al., 1994). Interpretation of the results of such surveys must take into account the extent to which segments of the community outside the health department contribute to meeting overall public health needs. Additional measures that may be of interest to communities are (1) per capita expenditures, by county and state, for public health activities; and (2) public health expenditures as a percentage of total health care expenditures, by state. Estimates for fiscal year 1993 indicate that national spending for core public health functions amounted to $11.4 billion, or 1.3 percent of total health expenditures in the United States (Public Health Foundation, 1994). Benchmarked rate of discretionary surgery. Small area variation studies have long been used to identify areas for which high rates reflect likely overuse of discretionary procedures such as coronary artery bypass surgery, cholecystectomy, cesarean section, or surgery for back pain. Benchmarked rate of hospital and intensive care unit (ICU) utilization . The use of acute care hospitals to treat many acute and chronic conditions is increasingly questioned in terms of both the risks associated with hospitalization and the high costs of care in

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring that setting. If possible, this indicator should be measured at the population rather than the institutional level. In addition, communities may want to examine the utilization of hospitals and ICUs for specific diagnoses. Underuse of generic pharmaceutical agents. The growing availability of computerized pharmacy records will make it possible to monitor the proportion of eligible prescriptions that are written for generic drugs of generally equal effectiveness but substantially lower cost (Walzak et al., 1994). Benchmarked supply of hospital beds, specialists, and primary care providers. Existing data from the American Hospital Association and physician organizations such as the American Medical Association (AMA) and the American Osteopathic Association can be used to determine the local supply of beds and physicians, while Medicare data or a state's hospital discharge data set can be used to adjust for border crossing and differences in age and sex across areas. A recently published analysis by researchers at Dartmouth provides such data for 1993 for all regions of the United States (Wennberg, 1996). Such analyses provide a potential source of data for benchmarking the level of personal health care resources within a community to numerous other communities that may achieve similar health outcomes with fewer resources. SAMPLE SET OF INDICATORS The 13 measures listed below represent a set of sample indicators that can be compiled by many communities throughout the United States: Percentage of inpatients age 65 and over who die without an advance directive. This measure would require a review of selected medical charts. The existence of an advance directive should be noted in the chart of patients with terminal conditions. Percentage of health plan enrollees who smoke. This measure would require a review of medical charts as well. Surveys at the community level may exist or could be developed.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Number of hospitals with a low volume of coronary artery bypass graft (CABG) surgery. Data for this indicator might be obtained from a state's hospital discharge data system or from Medicare databases. Number of infants weighing less than 1,500 grams born in hospitals without an advanced care nursery. These data are available from vital statistics, through the state or local health agency. Number of inpatient hospitalizations for asthma and diabetes. These data are available from hospitals or from a state's hospital discharge data system. Percentage of births without first trimester prenatal care. Information about prenatal care is available on birth certificates and from vital statistics at the state or local health department. Percentage of individuals without a usual source of care. This information is collected through the state-level Behavioral Risk Factor Surveillance System. Surveys at the community level may exist or could be developed. Performance ratio for 10 essential public health practices. This indicator requires a survey of "public health" providers, including state and local government agencies, voluntary non-profit community agencies, hospitals, physicians, clinics, community and migrant health centers, universities, federal agencies, foundations, and others. A survey protocol is available for use by communities (Miller et al., 1994). Ratio of discretionary surgeries to benchmark rates. This information can be compiled from hospital discharge or Medicare data. Communities also may consider conducting surveys. Percentage of prescriptions for generic drugs, by class of drug. Computerized pharmacy records can provide information, once the community has identified agents for which substitution should

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring be possible and has specified brand names that fall within the class. Ratio of hospital beds to benchmark rates. The American Hospital Association compiles this information. Ratio of medical specialists to benchmark rates. The local medical community and the AMA can provide these data. Ratio of primary care physicians to benchmark rates. Again, the local medical community or the AMA is a source of data. The proposed indicator set includes measures of individual behavior (advance directives and smoking), physical environment (low-volume CABG and obstetrical hospitals, avoidable ACS hospitalizations), and health care (missed opportunities, waste, and excess capacity). Most of the measures are tied to hospital data, reflecting its availability and reliability. One measure of public health efficiency (indicator 8) is included. Communities may also want to look for ways in which to incorporate measures of process efficiency and outcome efficiency. For example, communities may want to conduct surveys to learn about such process measures as whether purchasers of health care services are organized and whether consumers have adequate information. REFERENCES Baker, E.L., Melton, R.J., Stange, P.V., et al. 1994. Health Reform and the Health of the Public: Forging Community Health Partnerships. Journal of the American Medical Association 272:1276–1282. Bunker, J.P, Frazier, H.S., and Mosteller, F. 1995. The Role of Medical Care in Determining Health: Creating an Inventory of Benefits. In Society and Health. B.C. Amick, S. Levine, A.R. Tarlov, and D.C. Walsh, eds. New York: Oxford University Press. Emanuel, E.F. 1996. Cost Savings at the End of Life. What Do the Data Show? Journal of the American Medical Association 275:1907–1914. Fisher, E.S., Welch, H.G., and Wennberg, J.E. 1992. Prioritizing Oregon's Hospital Resources: An Example Based on Variations in Discretionary Medical Utilization. Journal of the American Medical Association 267:1925–1931. Guadagnoli, E., Hauptman, P.J., Ayanian, J.Z., Pashos, C.L., McNeil, B.J., and Cleary, P.D. 1995. Variation in the Use of Cardiac Procedures after Acute Myocardial Infarction. New England Journal of Medicine 333:573–578.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Hannan, E.L., Kilburn, H., Bernard, H., O'Donnell, J.F., Lukacik, G., and Shields, E.P. 1991. Coronary Artery Bypass Surgery: The Relationship between In-hospital Mortality Rate and Surgical Volume after Controlling for Clinical Risk Factors. Medical Care 29:1094–1107. IOM (Institute of Medicine). 1993. Access to Health Care in America . M. Millman, ed. Washington, D.C.: National Academy Press. Kaufman, A., and Waterman, R.E., eds. 1993. Health of the Public: A Challenge to Academic Health Centers. Strategies for Reorienting Academic Health Centers Toward Community Health Needs. San Francisco: University of California, Health of the Public Program. Luft, H.S., Garnick, D.W., Mark, D.H., and McPhee, S.J. 1990. Hospital Volume, Physician Volume, and Patient Outcomes: Assessing the Evidence . Ann Arbor, Mich.: Health Administration Press. McGinnis, J.M., and Foege, W.H. 1993. Actual Causes of Death in the United States. Journal of the American Medical Association 270:2207–2212. Miller, C.A., Moore, K.S., Richards, T.B., and Monk, J.D. 1994. A Proposed Method for Assessing the Performance of Local Public Health Functions and Practices. American Journal of Public Health 84:1743–1749. NCQA (National Committee for Quality Assurance). 1993. Health Plan Employer Data and Information Set and User's Manual, Version 2.0 (HEDIS 2.0). Washington, D.C.: NCQA. Park, R.E., Fink, A., Brook, R.H., et al. 1989. Physician Ratings of Appropriate Indications for Three Procedures: Theoretical Indications vs. Indications Used in Practice. American Journal of Public Health 79:445–447. Pennsylvania Health Care Cost Containment Council. 1995. A Consumer Guide to Coronary Artery Bypass Graft Surgery. Volume IV, 1993 Data. Pennsylvania's Declaration of Health Care Information. Harrisburg: Pennsylvania Health Care Cost Containment Council. Public Health Foundation. 1994. Measuring State Expenditures for Core Public Health Functions. Final report to the U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. Washington, D.C.: Public Health Foundation. Showstack, J., Lurie, N., Leatherman, S., Fisher, E., and Inui, T. 1996. Health of the Public: The Private-Sector Challenge. Journal of the American Medical Association 276:1071–1074. Summer, L. 1994. The Escalating Number of Uninsured in the United States. International Journal of Health Services 24(3):409–413. Turnock, B.J., Handler, A., Hall, W., Potsic, S., Nalluri, R., and Vaughn, E.H. 1994. Local Health Department Effectiveness in Addressing the Core Functions of Public Health. Public Health Reports 109:653–658. USDHHS (U.S. Department of Health and Human Services). 1991. Healthy People 2000: National Health Promotion and Disease Prevention Objectives . DHHS Pub. No. (PHS) 91-50212. Washington, D.C.: Office of the Assistant Secretary for Health. Walzak, D., Swindells, S., and Bhardwaj, A. 1994. Primary Care Physicians and the Cost of Drugs: A Study of Prescribing Practices Based on Recognition and Information Sources. Journal of Clinical Pharmacology 34:1159–1163. Wasson, J., Fleming, C., Bruskewitz, R., et al. 1993. The Treatment of Localized Prostate Cancer: What Are We Doing, What Do We Know, and What Should We Be Doing? Seminars in Urology 11(1):23–26.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring Weeks, W.B., Kofoed, L.L., Wallace, A.E., and Welch, H.G. 1994. Advance Directives and the Cost of Terminal Hospitalization. Archives of Internal Medicine 154:2077–2083. Welch, W.P., Verrilli, D., Washington, D.C., Katz, S.J., and Latimer E. 1996. A Detailed Description of Physician Services for the Elderly in the United States and Canada. Journal of the American Medical Association 275:1410–1416. Wennberg, J., ed. 1996. The Dartmouth Atlas of Health Care. Chicago: American Hospital Press.

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring TABLE A.5-1 Field Model Mapping for Sample Indicator Set: Health Care Resource Allocation Field Model Domain Construct Sample Indicators Data Sources Disease, Health and Function, Prosperity, Well-being   See community health profile Use to select benchmark communitiesa Individual Behavior Use of advance directives Percentage of inpatients 65 and over who die without an advance directiveb Chart review of selected records   Health risk behavior Percentage of health plan enrollees who smoke Chart review, HEDIS Physical Environment Inadequate regionalization of surgery Number of hospitals with low volume of CABG surgery Statewide hospital discharge database, Medicare databases   Inadequate regionalization of high-risk care Number of infants weighing 81,500 grams born in hospitals without an advanced care nursery Vital statistics   Avoidable hospitalizations Number of inpatient hospitalizations for asthma and diabetes Statewide hospital discharge database, hospital records Health Care Underprovision of basic services Percentage of births without first trimester care Vital statistics, state health department     Percentage of individuals without a usual source of care BRFSS

OCR for page 262
Improving Health in the Community: A Role for Performance Monitoring     Performance ratio for 10 essential public health practices Survey needed   Overprovision of services or waste Ratio of discretionary surgeries to benchmark ratesc Statewide hospital discharge database, Medicare databases     Percentage of prescriptions for generic drugs, by class of drugd Computerized pharmacy records   Excess health care capacity Ratio of beds to benchmark rates AHA data     Ratio of medical specialists to benchmark rates AMA, AOA, or local data     Ratio of primary care physicians to benchmark rates AMA, AOA, or local survey NOTE: AHA, American Hospital Association; AMA, American Medical Association; AOA, American Osteopathic Association; BRFSS, Behavioral Risk Factor Surveillance System; HEDIS, Health Plan Employer Data and Information Set. a Many of the measures in this indicator set are most easily interpreted by cross-sectional or longitudinal comparisons. Cross-sectional comparisons between one community and a similar benchmark community that appears to have equal resources (income) and outcomes (mortality), yet a lower-cost health system, represent one possibility. Longitudinal comparisons would allow individual communities to aim for improvement, regardless of their baseline rates on these numbers. b Evidence that advance directives are effective in themselves at reducing expenditures does not exist. They may, however, be a necessary precondition to other cost-saving interventions. c Surgical procedures prescribed by physicians in the absence of adequate provision of information to patients (potentially unwanted surgeries) include CABG, carotid endarterectomy, cholecystectomy, and back surgery. Small area analysis is only an indirect measure. Direct measure requires survey of information provision systems in a community. d Development of this measure will require (1) identifying agents for which substitution should be possible, (2) defining brand names that fall within that class, and (3) obtaining computerized pharmacy records for analysis.