Why Not the Best? Results from the National Scorecard on U.S. Health System Performance, 2008 T h e C o m m o n w e a lt h F u n d C o m m i s s i o n o n a H i g h P e r f o r m a n c e H e a lt h S y s t e m J ULY 2 0 0 8 t h e c o m m o n w e a lt h fu n d c o m m i s s i o n o n a h i g h p e r fo r m a n c e h e a lt h s ys t e m Membership James J. Mongan, M.D. Glenn M. Hackbarth, J.D. Mary K. Wakefield, Ph.D., R.N. Chair of the Commission Chairman Associate Dean President and CEO MedPAC School of Medicine Partners HealthCare System, Inc. Health Sciences Director and Professor George C. Halvorson Center for Rural Health Maureen Bisognano Chairman and CEO University of North Dakota Executive Vice President & COO Kaiser Foundation Health Plan, Inc. Institute for Healthcare Improvement Alan R. Weil, J.D. Robert M. Hayes, J.D. Executive Director Christine K. Cassel, M.D. President National Academy for State Health Policy President and CEO Medicare Rights Center President American Board of Internal Medicine Center for Health Policy Development and ABIM Foundation Cleve L. Killingsworth Chairman and CEO Steve Wetzell Michael Chernew, Ph.D. Blue Cross Blue Shield of Massachusetts Vice President Professor HR Policy Association Department of Health Care Policy Sheila T. Leatherman Harvard Medical School Research Professor School of Public Health Patricia Gabow, M.D. University of North Carolina Stephen C. Schoenbaum, M.D. CEO and Medical Director Judge Institute Executive Director Denver Health University of Cambridge Executive Vice President for Programs The Commonwealth Fund Robert Galvin, M.D. Gregory P. Poulsen Director, Global Health Senior Vice President Anne K. Gauthier General Electric Company Intermountain Health Care Deputy Director Assistant Vice President Fernando A. Guerra, M.D. Dallas L. Salisbury The Commonwealth Fund Director of Health President & CEO San Antonio Metropolitan Health District Employee Benefit Research Institute Cathy Schoen Research Director Sandra Shewry Senior Vice President for Director Research and Evaluation California Department of Health Services The Commonwealth Fund Glenn D. Steele, Jr., M.D., Ph.D. Rachel Nuzum President and CEO Senior Policy Director Geisinger Health System The Commonwealth Fund Allison Frey Program Associate The Commonwealth Fund t h e c o m m o n w e a lt h fu n d The Commonwealth Fund, among the first private foundations started by a woman philanthropist—Anna M. Harkness—was established in 1918 with the broad charge to enhance the common good. The mission of The Commonwealth Fund is to promote a high performing health care system that achieves better access, improved quality, and greater efficiency, particularly for society’s most vulnerable, including low-income people, the uninsured, minority Americans, young children, and elderly adults. The Fund carries out this mandate by supporting independent research on health care issues and making grants to improve health care practice and policy. An C OV ER P HO T O international program in health policy is designed to stimulate innovative policies © Jose Luis Pel aez , Inc . / and practices in the United States and other industrialized countries. B l e n d I m ag e s / C o r b i s Why Not the Best? RESULTS FROM THE NATIONAL SCOREC ARD ON U. S. HEALTH SYSTEM P ERFORMAN CE , 20 08 T he C ommon w ea lth F und C ommission on a H igh P erformance H ea lth S ystem J U LY 2 0 0 8 A BSTR AC T: Prepared for the Commonwealth Fund Commission on a High Performance Health System, the National Scorecard on U.S. Health System Performance, 2008, updates the 2006 Scorecard, the first comprehensive means of measuring and monitoring health care outcomes, quality, access, efficiency, and equity in the United States. The 2008 Scorecard, which presents trends for each dimension of health system performance and for individual indicators, confirms that the U.S. health system continues to fall far short of what is attainable, especially given the resources invested. Across 37 core indicators of performance, the U.S. achieves an overall score of 65 out of a possible 100 when comparing national averages with U.S. and international performance benchmarks. Overall, performance did not improve from 2006 to 2008. Access to health care significantly declined, while health system efficiency remained low. Quality metrics that have been the focus of national campaigns or public reporting efforts did show gains. Support for this research was provided by The Commonwealth Fund. This and other Fund publications are available online at www.commonwealthfund.org. To learn more about new publications when they become available, visit the Fund’s Web site and register to receive e-mail alerts. Commonwealth Fund pub. no. 1150. P hoto : Larry M u lvehi l l / C orbis Contents Preface 5 Acknowledgments 6 List of Exhibits 7 Executive Summary 9 Introduction 15 The Scorecard: Measuring and Monitoring Health System Performance 17 Findings from the 2008 National Scorecard 18 OVERALL SCORES AND TRENDS 18 HEALTHY LIVES 18 QUALIT Y OF CARE 20 HEALTH CARE ACCESS 28 EFFICIENCY OF THE HEALTH SYSTEM 31 EQUIT Y IN THE HEALTH SYSTEM 35 SYSTEM CAPACIT Y TO INNOVATE AND IMPROVE 38 Summary and Implications 40 Notes 43 Appendices 47 Further Reading 60 P hoto : B l end I mages , LLC Preface As Chairman and Executive Director of the Commonwealth Although the task of moving to a system that is Fund Commission on a High Performance Health truly high performing is enormous, the stakes are even System, we are pleased to introduce the findings from the higher if we fail. The Commission’s National Scorecard Commission’s National Scorecard on U.S. Health System offers targets for change. The Scorecard underscores the Performance, 2008. Now in its second edition, the 2008 need for new national policies that pursue coverage and report presents current information and trends on the improvements in quality and efficiency simultaneously. nation’s progress toward achieving a system of care that It is essential to start as soon as possible to realize the affords better access, higher quality, and greater efficiency potential of accumulating substantial gains over time. for everyone. The December 2007 report, Bending the Curve: Options In September 2006, the Commission issued the first for Achieving Savings and Improving Value in U.S. Health National Scorecard as a means of setting realistic targets Spending, indicates it would be possible to save $1.5 trillion and monitoring change over time across a broad array of in national health expenditures over the next decade and indicators of health system performance spanning healthy improve the value of health care in the U.S., if aggressive lives, quality, access, efficiency, and equity. The first efforts start now. assessment revealed substantial room for improvement With the upcoming 2008 presidential election, there is across all dimensions. Despite many pockets of excellence, a window of opportunity to transform our health system overall the U.S. performs far below what is achievable. to one that gives everyone the chance to lead longer, This 2008 update of the National Scorecard shows that healthier, and more productive lives. In its report, A the nation continues to exhibit suboptimal performance High Performance Health System for the United States: An relative to benchmarks. Despite high and rising health Ambitious Agenda for the Next President, the Commission care expenditures, the U.S. is actually losing ground in recommended five strategies for health reform that must providing access to care. Health care quality remains highly be pursued together to move the nation in the right dependent on where you live and whom you see for care, direction. We hope to see serious discourse and bold which is inconsistent with the idea that all Americans action—enriched by these findings from the National receive the same high-quality care. At the same time, we Scorecard—begin in earnest next year. can begin to see what is possible when there is appropriate James J. Mongan, M.D.Stephen C. Schoenbaum, M.D. leadership and concerted efforts to set standards of Chairman Executive Director performance and ensure that improvement occurs. The Commonwealth Fund Commission on a High Performance Health System 5 Acknowledgments Special thanks go to Cathy Schoen, M.S., senior vice University of Minnesota School of Public Health; Sir president of The Commonwealth Fund, for working with Brian Jarman, M.D., Imperial College, London, U.K.; the Commission on a High Performance Health System to Ashish Jha, M.D., M.P.H., and Arnold Epstein, M.D., conceptualize and oversee the development and updating Harvard School of Public Health; Jeffrey Linder, M.D., of the Scorecard, and to Sabrina K. H. How, M.P.A., senior M.P.H., Brigham and Women’s Hospital; J. Michael research associate for The Commonwealth Fund, and McWilliams, M.D., Harvard Medical School; Vincent Douglas McCarthy, M.B.A., senior research adviser for Mor, Ph.D., Brown University; Deirdre Mylod, Ph.D., The Commonwealth Fund, for research, writing, and and Suzanne Coshow, Ph.D., Press Ganey Associates, preparation of the Scorecard and related materials. Inc.; Ellen Nolte, Ph.D., and C. Martin McKee, M.D., Five members of the Commonwealth Fund London School of Hygiene and Tropical Medicine; Commission on a High Performance Health System Michael Pineau and the patient safety team at Qualidigm; worked along with senior Fund staff to review and and Chunliu Zhan, M.D., Ph.D., Agency for Healthcare select indicators and design the initial Scorecard. These Research and Quality (AHRQ). Bisundev Mahato, include: Maureen Bisognano, executive vice president Columbia University Mailman School of Public Health, and COO, Institute for Healthcare Improvement; and Dina Belloff, M.A., Rutgers Center for State Health Michael Chernew, Ph.D., professor, Harvard Medical Policy, provided programming and analytical support. School; George Halvorson, chairman and CEO, Kaiser Other experts provided assistance with data updates. Foundation Health Plan, Inc.; Sheila Leatherman, We thank Karen Ho, M.H.S., and Jeff Brady, M.D., M.P.H., research professor, University of North Carolina; and at AHRQ; David Hunt, M.D., and Rebecca Kliman, M.D., Alan Weil, J.D., M.P.P., executive director, National at the Centers for Medicare and Medicaid Services; Alan Academy for State Health Policy. Simon, M.D., at the National Center for Health Statistics; The Commission wishes to thank the researchers Dale Shaller, M.P.A., and the AHRQ Consumer Assessment who helped develop indicators and conducted data of Healthcare Providers and Systems (CAHPS) Database analyses to update the Scorecard and accompanying team; and Jeff Van Ness and Joachim Buess at the National chartpack. These include: Gerard Anderson, Ph.D., and Committee for Quality Assurance. Robert Herbert, Johns Hopkins Bloomberg School of Additionally, we thank the following Commonwealth Public Health; Peter Cunningham, Ph.D., Center for Fund staff: Karen Davis, Ph.D., and Steve Schoenbaum, Studying Health System Change; Elliott Fisher, M.D., M.D., for reviewing drafts; Martha Hostetter, Chris M.P.H., Jason Sutherland, Ph.D., and David Radley, Hollander, and Paul Frame for editing; and Jim Walden M.P.H., Dartmouth Medical School; Leslie Grant, Ph.D., of Walden Creative for graphic design. 6 List of Exhibits E X H I BIT 1Scores: Dimensions of a High Performance Health System E X H I BIT 2National Scorecard on U.S. Health System Performance, 2008: Scores on 37 Key Performance Indicators E X H I BIT 3International Comparison of Spending on Health, 1980–2005 Healthy Lives E X H I BIT 4Mortality Amenable to Health Care Quality E X H I BIT 5Receipt of Recommended Screening and Preventive Care for Adults E X H I BIT 6 Chronic Disease Under Control: Diabetes and Hypertension E X H I BIT 7 Hospitals: Quality of Care for Heart Attack, Heart Failure, and Pneumonia E X H I BIT 8 Transition Care: Hospital Discharge and Follow-Up Care for Chronically Ill Patients E X H I BIT 9 Nursing Homes: Hospital Admission and Readmission Rates Among Nursing Home Residents E X H I BIT 10Hospital-Standardized Mortality Ratios E X H I BIT 11Difficulty Getting Care on Nights, Weekends, Holidays Without Going to the Emergency Room, Among Sicker Adults E X H I BIT 12 Patient-Centered Hospital Care: Staff Managed Pain, Responded When Needed Help, and Explained Medicines, by Hospitals, 2007 Access E X H I BIT 13 Percent of Adults Ages 18–64 Uninsured by State E X H I BIT 14Uninsured and Underinsured Adults, 2007 Compared with 2003 E X H I BIT 15Medical Bill Problems or Medical Debt Efficiency E X H I BIT 16 Test Results or Medical Records Not Available at Time of Appointment, Among Sicker Adults E X H I BIT 17Medicare Hospital 30-Day Readmission Rates E X H I BIT 18 Quality and Costs of Care for Medicare Patients Hospitalized for Heart Attacks, Hip Fractures, or Colon Cancer, by Hospital Referral Regions, 2004 E X H I BIT 19 Percentage of National Health Expenditures Spent on Insurance Administration, 2005 E X H I BIT 20 Physicians’ Use of Electronic Medical Records Equity E X H I BIT 21Equity: Ratio Scores for Insurance, Income, and Race/Ethnicity E X H I BIT 22Untreated Dental Caries, by Age, Race/Ethnicity, and Income, 2001–2004 E X H I BIT 23Ambulatory Care–Sensitive (Potentially Preventable) Hospital Admissions, by Race/Ethnicity and Patient Income Area, 2004/2005 7 P hoto : R andy H adaway Executive Summary Every family wants the best care for an ill or injured family health outcomes achieved by the leading countries. The member. Most are grateful for the care and attention U.S. now ranks last out of 19 countries on a measure of received. Yet, evidence in the National Scorecard on U.S. mortality amenable to medical care, falling from 15th as Health System Performance, 2008, shows that care typically other countries raised the bar on performance. Up to falls far short of what is achievable. Quality of care is highly 101,000 fewer people would die prematurely if the U.S. variable, and opportunities are routinely missed to prevent could achieve leading, benchmark country rates. disease, disability, hospitalization, and mortality. Across The exception to this overall trend occurred for 37 indicators of performance, the U.S. achieves an overall quality metrics that have been the focus of national score of 65 out of a possible 100 when comparing national campaigns or public reporting. For example, a key patient averages with benchmarks of best performance achieved safety measure—hospital standardized mortality ratios internationally and within the United States. (HSMRs)—improved by 19 percent from 2000–2002 Even more troubling, the U.S. health system is on the to 2004–2006. This sustained improvement followed wrong track. Overall, performance has not improved since widespread availability of risk-adjusted measures coupled the first National Scorecard was issued in 2006. Of greatest with several high-profile local and national programs to concern, access to health care has significantly declined. improve hospital safety and reduce mortality. Hospitals As of 2007, more than 75 million adults—42 percent of are showing measurable improvement on basic treatment all adults ages 19 to 64—were either uninsured during guidelines for which data are collected and reported the year or underinsured, up from 35 percent in 2003. At nationally on federal Web sites. Rates of control of two the same time, the U.S. failed to keep pace with gains in common chronic conditions, diabetes and high blood EXHIBIT 1 Scores: Dimensions of a High Performance Health System Healthy Lives 75 2006 Revised 72 2008 Quality 72 71 Access 67 58 Efficiency 52 53 Equity 70 71 67 OV E R A L L S C O RE 65 0 100 Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 9 pressure, have also improved significantly. These measures exhibited no change (or were not updated). Exhibit 2 lists are publicly reported by health plans, and physician indicators and summarizes scores and benchmark rates. groups are increasingly rewarded for results in improving As observed in the first Scorecard, the bottom group treatment of these conditions. of hospitals, health plans, or geographic regions is often The U.S. spends twice per capita what other major well behind even average rates, with as much as a fivefold industrialized countries spend on health care, and costs spread between top and bottom rates. On key indicators, continue to rise faster than income. We are headed toward a 50 percent improvement or more would be required to $1 of every $5 of national income going toward health care. achieve benchmark levels. We should expect a better return on this investment. Performance on measures of health system efficiency SCOR ECA R D H IGH LIGHTS remains especially low, with the U.S. scoring 53 out of 100 on A N D K E Y FI N DI NGS measures gauging inappropriate, wasteful, or fragmented care; avoidable hospitalizations; variation in quality The U.S. continues to perform far below what is and costs; administrative costs; and use of information achievable, with wide gaps between average and technology. Lowering insurance administrative costs benchmark performance across dimensions. Despite alone could save up to $100 billion a year at the lowest some encouraging pockets of improvement, the country rates. country as a whole has failed to keep pace with levels National leadership is urgently needed to yield greater of performance attained by leading nations, delivery value for the resources devoted to health care. systems, states, and regions. Following are major highlights from the Scorecard T H E NATIONA L SCOR ECA R D by performance dimension: The National Scorecard includes 37 indicators in five HEAL T HY LI V ES : A V ERAGE S C ORE 7 2 dimensions of health system performance: healthy •• Preventable mortality: The U.S. fell to last place among lives, quality, access, efficiency, and equity. U.S. average 19 industrialized nations on mortality amenable to performance is compared with benchmarks drawn health care—deaths that might have been prevented from the top 10 percent of U.S. states, regions, health with timely and effective care. Although the U.S. plans, hospitals, or other providers or top-performing rate improved by 4 percent between 1997–1998 and countries, with a maximum possible score of 100. If 2002–2003 (from 115 to 110 deaths per 100,000), rates average U.S. performance came close to the top rates improved by 16 percent on average in other nations, achieved at home or internationally, then average scores leaving the U.S. further behind. would approach 100. •• Activity limitations: More than one of every six In 2008, the U.S. as a whole scored only 65, compared working-age adults (18%) reported being unable to with a score of 67 in 2006—well below the achievable work or carry out everyday activities because of health benchmarks (Exhibit 1).* Average scores on each of the problems in 2006—up from 15 percent in 2004. This five dimensions ranged from a low of 53 for efficiency to increase points to the need for better prevention and 72 for healthy lives. management of chronic diseases to enhance quality of On those indicators for which trend data exist, life and capacity to work, especially among younger performance compared with benchmarks more often adults as they age. worsened than improved, primarily because of declines in national rates between the 2006 and 2008 Scorecards. Q UALI T Y : A V ERAGE S C ORE 7 1 Overall, national scores declined for 41 percent of •• Effective care: Control of diabetes and high blood indicators, while one-third (35%) improved, and the rest pressure improved markedly from 1999–2000 to 2003–2004 for adults, according to physical exams *The overall score for 2006 changed from 66 to 67 due to revisions conducted on a nationally representative sample. in baseline data and substitution of top U.S. states for countries as Among adults with diabetes, rates of at least fair the benchmark for infant mortality. See methodology box on p. 17 for further details. control of blood sugar increased from 79 percent to 10 EXHIBIT 2 national scorecard on u.s. health system Performance, 2008: scores on 37 Key Performance indicators u.s. national rate 2008 score: 2006 2008 Benchmark ratio of u.s. to indicator scorecard scorecard Benchmark rate Benchmark oV e r a l l s C o r e 65 h e a lT h y l i V e s Top 3 of 1 Mortality amenable to health care, deaths per 100,000 population 115 110 69 63 19 countries 2 Infant mortality, deaths per 1,000 live births 7.0 6.8 Top 10% states 4.7 69 3 Healthy life expectancy at age 60, years Various * Various Various 87* Adults under 65 limited in any activities because of 4 14.9 17.5 Top 10% states 11.5 66 physical, mental, or emotional problems, % 5 Children missed 11 or more school days due to illness or injury, % 5.2 * Top 10% states 3.8 73* Q ua l i T y 6 Adults received recommended screening and preventive care, % 49 50 Target 80 62 7 Children received recommended immunizations and preventive care Various Various Various Various 86 8 Needed mental health care and received treatment Various Various Various Various 76 9 Chronic disease under control Various Various Various Various 76 Hospitalized patients received recommended care for heart 10 84 90 Top hospitals 100 90 att ack, heart failure, and pneumonia (composite), % 65+ yrs, 11 Adults under 65 with accessible primary care provider, % 66 65 85 76 High income 12 Children with a medical home, % 46 * Top 10% states 60 77* 13 Care coordination at hospital discharge Various Various Various Various 74 14 Nursing homes: hospital admissions and readmissions Various Various Various Various 65 Top 25% 15 Home health: hospital admissions, % 28 28 17 62 agencies Best of 16 Patient reported medical, medication, or lab test error, % 34 32 19 59 7 countries 17 Unsafe drug use Various Various Various Various 55 18 Nursing home residents with pressure sores Various Various Various Various 66 Top 10% 19 Hospital-standardized mortality ratios, actual to expected deaths 101 82 74 90 hospitals Best of 20 Ability to see doctor same/next day when sick or need medical care % 47 46 81 57 6 countries Very/somewhat easy to get care after hours without Best of 21 38 25 72 35 going to the emergency room, % 6 countries Doctor-patient communication: always listened, explained, 90th %ile 22 54 57 75 75 showed respect, spent enough time, % health plans Best of 23 Adults with chronic conditions given self-management plan, % 58 * 65 89* 6 countries 24 Patient-centered hospital care Various Various Various Various 87 aC C e s s 25 Adults under 65 insured all year, not underinsured, % 65 58 Target 100 58 Best of 26 Adults with no access problem due to costs, % 60 63 95 66 7 countries Families spending <10% of income or <5% of income, if low 27 81 77 Target 100 77 income, on out-of-pocket medical costs and premiums, % Population under 65 living in states where premiums for employer- 28 58 25 Target 100 25 sponsored coverage are <15% of median household income, % 29 Adults under 65 with no medical bill problems or medical debt, % 66 59 Target 100 59 eFFiCienCy 30 Potential overuse or waste Various Various Various Various 41 went to emergency room for condition that could Best of 31 26 21 6 29 have been treated by regular doctor, % 7 countries 32 Hospital admissions for ambulatory care–sensitive conditions Various Various Various Various 56 33 Medicare hospital 30-day readmission rates, % 18 18 10th %ile regions 14 76 Medicare annual costs of care and mortality for heart att acks, hip fractures, $26,829 $28,011 $24,906 34 10th %ile regions 89 or colon cancer (annual Medicare outlays; deaths per 100 benefi ciaries) 30 30 27 35 Medicare annual costs for chronic diseases: Diabetes, heart failure, COPD Various Various Various Various 71 Top 3 of 36 Health insurance administration as percent of national health expenditures 7.4 7.5 2.3 31 11 countries Best of 37 Physicians using electronic medical records, % 17 28 98 29 7 countries Various = indicators that comprise two or more related measures; scores average the individual ratios for each component. COPD = chronic obstructive pulmonary disease. * Indicator not updated; baseline score same as 2006. See Exhibit 21 on page 35 for Equity scores; see Appendices A and B for more details on data and sources. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 11 88 percent from 1999–2000 to 2003–2004. Among U.S. adults reported going without needed care adults with hypertension, rates of control of high because of costs in 2007, versus only 5 percent in blood pressure increased from 31 percent to 41 percent the benchmark country. over the same time period. Yet, a 30 to 60 percentage •• Affordable care: As insurance premiums rose faster point difference remains between top- and bottom- than wages, the share of nonelderly adults living performing health plans. Hospitals’ adherence to in a state where group health insurance premiums treatment standards for heart attack, heart failure, averaged less than 15 percent of household income and pneumonia also improved from 2004 to 2006, dropped sharply, from 58 percent in 2003 to 25 percent but with a persistent gap between leading and lagging in 2005. By 2007, two of five adults (41%) reported hospital groups. Delivery rates for basic preventive they had medical debt or problems with medical bills, care failed to improve: as of 2005, only half of adults up from 34 percent in 2005. received all recommended preventive care. •• Coordinated care: Heart failure patients were more E F F I C IEN C Y : A V ERAGE S C ORE 5 3 likely to receive hospital discharge instructions in •• Inappropriate, wasteful, or fragmented care: In 2007, as 2006 (68%) than in 2004 (50%), but rates varied in 2005, U.S. patients were much more likely—three widely between top and bottom hospital groups to four times the benchmark rate—than patients in (from 94% to 36%). Hospitalizations increased other countries to report having had duplicate tests or among nursing home residents from 2000 to 2004, that medical records or test results were not available as did rehospitalizations for patients discharged at the time of their appointment. to skilled nursing facilities—signaling a need to •• Avoidable hospitalizations: Average rates of hospital improve long-term care and transitions between readmissions within 30 days remained high, at 18 health care providers. percent in both 2003 and 2005. Rates in the highest •• Safe care: One key indicator of patient safety— regions were 50 percent higher than in the lowest hospital standardized mortality ratios—improved regions. Rates of hospitalizations for preventable significantly since the first Scorecard, with a 19 conditions decreased somewhat from 2002–2003 to percent decline. Safety risks, however, remain 2004–2005, but continued to vary two- to fourfold high as one-third of adults with health problems across hospital regions and states. reported mistakes in their care in 2007. Drug safety •• Variation in quality and costs: Among Medicare is of particular concern. Rates of visits to physicians patients treated for heart attacks, hip fractures, or or emergency departments for adverse drug effects colon cancer, a high proportion of regions with the increased by one-third between 2001 and 2004. lowest mortality rates also had lower total costs, •• Patient-centered, timely care: In 2007, as in 2005, less than indicating that it is possible to save lives and lower half of U.S. adults with health problems were able to get costs through more effective, efficient systems. The a rapid appointment with a physician when they were total costs of caring for patients with chronic disease sick. They also were the most likely among adults in varied twofold across regions. seven countries surveyed to report difficulty obtaining •• Administrative costs : U.S. health insurance health care after hours without going to the emergency administrative costs as a share of total health department, and this rate increased from 61 percent to spending are 30 percent to 70 percent higher than 73 percent since 2005. Within the U.S., there is wide in countries with mixed private/public insurance variation among hospitals in terms of patient reports systems and three times higher than in countries of how well staff responded to their needs. with the lowest rates. •• Information systems: U.S. primary care physicians’ A C C ESS : A V ERAGE S C ORE 5 8 use of electronic medical records (EMRs) increased •• Insurance and access: As of 2007, 75 million from 17 percent to 28 percent from 2001 to 2006. working-age adults (42%) were either uninsured Still, the U.S. lags far behind leading countries, or underinsured, a sharp increase from 61 million where EMRs are now used by nearly all physicians (35%) in 2003. More than one-third (37%) of all (98%) to improve care. 12 E Q UI T Y : A V ERAGE S C ORE 7 1 national investment in research regarding clinical and •• Disparities: Compared with their white, higher- cost-effectiveness—what works well for which patients income, or insured counterparts, minorities, low- and when—has failed to keep pace to inform health care income, or uninsured adults and children were decision-making. generally more likely to wait when sick, to encounter delays and poorly coordinated care, and to have SU M M A RY A N D I M PLICATIONS untreated dental caries, uncontrolled chronic disease, avoidable hospitalizations, and worse outcomes. They P O T EN T IAL F OR IM P RO V EMEN T were also less likely to receive preventive care or have Overall, the National Scorecard on U.S. Health System an accessible source of primary care. Performance, 2008, finds that the U.S. is losing ground in •• Reducing gaps: Among blacks and Hispanics, it providing access to care and has uneven health care quality. would require a 19 percent to 25 percent decrease The Scorecard also finds broad evidence of inefficient in the risk of poor health outcomes and inadequate and inequitable care. Average U.S. performance would or inefficient care to reach parity with whites. Gaps have to improve by more than 50 percent across multiple for uninsured and low-income populations are still indicators to reach benchmark levels of performance. wider: it would require a 34 percent to 39 percent Closing performance gaps would bring real improvement on indicators of health care access, benefits in terms of health, patient experiences, and quality, and efficiency to achieve equity with insured savings. For example: and higher-income populations. •• Up to 101,000 fewer people would die prematurely each year from causes amenable to health care if the SYS T EM C A P A C I T Y T O INNO V A T E AND U.S. achieved the lower mortality rates of leading IM P RO V E : NO T S C ORED countries. The capacity to innovate and improve is fundamental to •• Thirty-seven million more adults would have an a high-performing health care system. It includes: accessible primary care provider, and 70 million more •• a care system that supports a skilled and motivated adults would receive all recommended preventive health care workforce, with an emphasis on primary care. care and population health; •• The Medicare program could potentially save at •• a culture of quality improvement and continuous least $12 billion a year by reducing readmissions learning that promotes and rewards recognition or by reducing hospitalizations for preventable of opportunities to reduce errors and improve conditions. outcomes; and •• Reducing health insurance administrative costs to the •• investment in public health initiatives, research, and average level of countries with mixed private/public information necessary to inform, guide, and drive insurance systems (Germany, the Netherlands, and health care decisions and improvement. Switzerland) would free up $51 billion, or more than On all three aspects, the U.S. currently under-invests in half the cost of providing comprehensive coverage to the capacity of the health system to innovate and improve. all the uninsured in the U.S. Reaching benchmarks U.S. payment systems undervalue primary care and fail to of the best countries would save an estimated $102 support providers’ efforts to manage and coordinate care. billion per year. Studies indicate that health care teams and well-organized Studies further document the cost in lives and lost work processes can achieve significant gains in quality productivity from the nation’s failure to provide secure and safety with more efficient use of resources. Yet, health health insurance to all. Based on areas within the U.S. professionals are rarely trained to work in teams, and larger that achieve superior outcomes at lower costs, it should organized delivery systems that employ multidisciplinary be possible to close gaps in health care quality and access, health professionals are not the norm. There is little and to reduce costs significantly. investment in spreading best practices, and incentives Several implications for policy emerge from the are rarely designed to reward or support improved quality Scorecard findings: and greater efficiency. In an era of rapid medical advances, 13 WHA T RE C EI V ES A T T EN T ION GE T S AIMING HIGHER IM P RO V ED The 2008 National Scorecard documents the human Notably, all of the quality indicators showing significant and economic costs of failing to address the problems in improvement have been targets of national and collaborative our health system. Recent analysis suggests it could be efforts to improve, informed by data with measurable possible to insure everyone and achieve significant savings benchmarks and indicators reached by consensus. with improved value over the next decade.3 Health care Conversely, there was failure to improve in areas such as expenditures are projected to double to $4 trillion, or 20 mental health care, primary care, hospital readmission rates, percent of national income, over the next decade, and or adverse drug events for which focused efforts to assess millions more U.S. residents are on a path to becoming and improve at the community or facility level are lacking. uninsured or underinsured, absent new policies. We Further, the continued failure to adopt interoperable health need to change directions, starting with the recognition information technology makes it difficult to generate the that access to care, health care quality, and efficiency are information necessary to document performance and interrelated. monitor improvement efforts. Aiming higher and moving on a more positive path will require strategies targeting the multiple sources of poor B E T T ER P RIMARY C ARE AND C ARE health system performance. These strategies include: C OORDINA T ION HOLD P O T EN T IAL F OR •• universal and well-designed coverage that ensures IM P RO V ED OU T C OMES A T LOWER C OS T S affordable access and continuity of care, with low Hospital readmission rates and rates of potentially administrative costs; preventable hospitalizations for ambulatory care–sensitive •• incentives aligned to promote higher quality and conditions remain high and variable across the country, as more efficient care; do total costs for the chronically ill. Studies indicate that it •• care that is designed and organized around the is possible to prevent hospitalization or rehospitalizations patient, not providers or insurers; with better primary care, discharge planning, and follow- •• widespread implementation of health information up care—an integrated, systems approach to care. technology with information exchange; Multiple indicators highlight the fact that the U.S. •• explicit national goals to meet and exceed benchmarks has a weak primary care foundation. Investing in primary and monitor performance; and care with enhanced capacity to provide patients with •• national policies that promote private–public round-the-clock access, manage chronic conditions, collaboration and high performance.4 and coordinate care will be key steps in moving to more Rising costs put families, businesses, and public organized care systems.1 budgets under stress, pulling down living standards for However, current payment incentives for hospitals, middle- as well as low-income families. New national physicians, and nursing homes do not support coordination policies that take a coherent, whole-system, population of care or efficient use of expensive, specialized care.2 view are essential for the nation’s future health and Information also fails to flow with patients across sites economic security. of care due to lack of health information technology and information exchange systems. These inefficiencies require innovative payment policies as well as care delivery approaches to improve outcomes for patients and use resources more efficiently. 14 Introduction In the first decade of the 21st century, the nation’s health industrialized countries spend—and has had rapid rates care system faces challenges on multiple fronts. The of cost growth over the past two decades (Exhibit 3). number of uninsured has increased by 8.6 million since Evidence continues to mount that the quality of care 2000, as employer-sponsored coverage continued to erode is uneven and often suboptimal.8 Quality encompasses even during a period of economic expansion. There were not only whether patients receive care that is safe and 47 million uninsured Americans as of 2006.5 Affordable scientifically proven, but also whether physicians insurance is of concern to families, employers, and public communicate well with patients and coordinate care programs: as health care costs continue to rise far faster effectively when patients transition from one place to than incomes, financial protection and access to care for another. Yet, providers’ financial incentives typically middle- as well as low-income families are increasingly encourage doing more rather than supporting high-quality, at risk.6 integrated care across settings, episodes, and conditions U.S. health expenditures, already the highest in the with more efficient use of resources. Too often, patients world, are projected to double and reach 20 percent of are left to cope with what is, in effect, a fragmented “non- the nation’s gross domestic product (GDP) by 2017, system” of care. Reflecting broad public concerns with with even higher shares of GDP going toward health access, costs, and care experiences, the percentage of care over the longer term.7 The United States spends the patients expressing dissatisfaction with the health care most per person on health care—twice what other major system doubled from 1998 to 2006.9 EXHIBIT 3 International Comparison of Spending on Health, 1980–2005 Average spending on health per capita ($US PPP*) Total expenditures on health as percent of GDP $7000 16% United States Germany 14 6000 Canada France 12 5000 Australia United Kingdom 10 4000 8 3000 6 United States 2000 Germany 4 Canada France 1000 2 Australia United Kingdom 0 0 80 82 84 86 80 88 82 90 84 92 86 94 88 96 90 98 92 00 94 02 96 04 98 00 02 04 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 19 20 19 20 19 20 20 20 * PPP=Purchasing Power Parity. Data: OECD Health Data 2007, Version 10/2007. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 15 Unlike virtually all other industrialized countries, the The National Scorecard on U.S. Health System U.S. fails to ensure universal coverage of its population. Performance, 2008, updates the analysis to assess This failing has serious consequences: poorer health from current performance and changes over time, based on lack of timely access to care; health conditions that, left the most recent data available. By contrasting national unchecked, become costlier to treat; premature death; performance with benchmarks, the Scorecard provides and reduced economic output from a less productive, targets for action and a yardstick against which to assess sicker workforce.10 Other nations spend less on health new policies over time. In the sections that follow, we care, achieve better health outcomes, and cover their describe how the Scorecard works and present overall entire populations. This indicates that the U.S. is not findings and results for five core dimensions of health getting high value commensurate with its investment in system performance. We conclude with an analysis of the health care system.11 cross-cutting themes and implications for national policy Developing policies to move the U.S. toward a higher- and improvement initiatives. value health system over time, and evaluating the effects of particular health policies relative to goals, requires a means to monitor health system performance across all of its dimensions. To meet this need for a whole-system view, the Commonwealth Fund Commission on a High Performance Health System created a National Scorecard on U.S. Health System Performance in 2006.12 Spanning healthy lives, quality, access, efficiency, and equity, the Scorecard found that U.S. health system performance fell far short of what should be attainable, based on benchmarks of achieved performance, and uncovered broad evidence of opportunities to improve. 16 The Scorecard: Measuring and Monitoring Health System Performance The National Scorecard on U.S. Health System Performance •• equity, which looks at disparities among population provides a unique, comprehensive approach to measuring groups in terms of health status, care, and coverage. and monitoring the performance of the nation’s health care The 2008 Scorecard uses the same framework, system. The Commonwealth Fund Commission on a High methods, and set of 37 performance indicators included in Performance Health System developed the Scorecard to the first Scorecard published in 2006. The analysis assesses serve three central goals: current performance as well as changes over time. •• to provide benchmarks for assessing health system For each indicator, the Scorecard compares national performance; performance against benchmark levels achieved by top- •• to have a mechanism for monitoring change over performing groups within the U.S. or other countries. In a time; and few instances, benchmarks reflect targets or policy goals. The •• to be able to estimate the effects of proposed policies report updates the benchmarks whenever top performance to improve performance. improved from baseline values observed in the 2006 report. The Scorecard includes key indicators of national health Each score is a simple ratio of the current U.S. average system performance organized into five core dimensions: performance to the benchmark representing best levels of •• healthy lives, which includes life expectancy, mortality, and achievement, with a maximum possible score of 100. prevalence of disability and limitations due to health; To examine trends, we compare the baseline and •• quality, a broad measure covering the extent to which current national averages as well as the change in the the care delivered is effective and well-coordinated, range of performance. Time trends typically capture two safe, timely, and patient-centered; years and up to five years for some indicators. Where •• access, which is concerned with participation in the indicators could not be updated, we retained baseline health care system and the affordability of insurance values to score. The tables in Appendix A present details coverage and medical services; for all indicators. (See box for further information on •• efficiency, which assesses overuse or inappropriate methodology.) An extensive Scorecard Chartpack is use of services, preventable hospitalizations and available online at www.commonwealthfund.org. readmissions, regional variation in quality and cost, Future editions of the Scorecard will continue to administrative complexity, and use of information monitor trends and add or improve indicators as new systems; and data become available. S C ORE C ARD ME THODOLOGY based on rates achieved by the top 10 percent would indicate a move in a positive direction, of U.S. states, regions, hospitals, health plans, we divided the national average by the The National Scorecard on U.S. Health System or other providers or top countries. Where benchmark. Where lower rates would indicate Performance, 2008, includes a set of 37 core patient data were available only at the national indicators that builds on metrics developed a positive direction (e.g., mortality, medical level, we identified benchmarks based on the errors), we divided the benchmark by the by public and private quality improvement experiences of high-income, insured individuals. efforts, as well as several unique indicators national average. Where updated data were Four access benchmarks aim for logical policy created for the Scorecard that are not currently not available, we retained baseline scores. goals, such as 100 percent of the population tracked elsewhere. to be adequately insured. For one quality To summarize, we averaged ratios within The 2008 Scorecard uses the same set of indicator—adults getting all recommended dimension and averaged dimensions for an indicators used in the 2006 Scorecard, with preventive care—we set a target rate of 80 overall score. For equity, we compared the one exception reflecting a change in the data percent, since rates even among high-income, percentage of the group at risk (e.g., percent source: a general measure of mental health care insured populations were low. not receiving recommended care, percent was replaced by a more specific measure of We updated benchmarks whenever they uninsured) by insurance, income, and race/ treatment of a major depressive episode. Many improved. Thus, it is possible for scores to ethnicity on a subset of indicators. We also of the indicators are composites that summarize decline if benchmarks improve faster than included a few specific indicators of health performance across multiple measures. Of the the national average. For costs, we used care equity to highlight areas of concern. The underlying 61 data elements, 53 were updated. the most recent data on the lowest-cost risk ratios compare rates for insured relative Almost all updates spanned at least two years; groups as benchmarks. For patient-reported to uninsured; high income to low income; and more than one-third assessed change over experiences in hospitals, we used the newly three to five years. For each indicator, we whites to blacks and Hispanics. available broad sample to benchmark, rather present national data for the baseline used in than the pilot set in the first Scorecard. For We recalculated baseline scores when the 2006 Scorecard and most recent year. infant mortality, we switched the benchmark necessary due to data revisions. As a result, Scoring consists of a simple ratio that from countries to top U.S. states to ensure the overall baseline score changed from 66 to compares national performance to the comparable indicator methods. 67 for the 2006 Scorecard. See Appendices A benchmark, with a maximum score of 100. To score, we calculated ratios of average and B for scoring tables and details regarding For each indicator, we identified benchmarks rates to the benchmark. Where higher rates indicator data, years, and sources. 17 Findings from the 2008 National Scorecard OVER ALL SCORES AND TRENDS: and consistently ranks poorly in comparison with other 2008 SCORECARD COMPARED WITH countries on measures of healthy lives, care experiences, 2006 SCORECARD and efficiency. The following sections summarize findings of the Overall, the National Scorecard on U.S. Health System 2008 Scorecard, highlighting individual indicators and Performance, 2008, finds that the U.S. health system changes in performance since the 2006 Scorecard. continues to perform far below benchmarks of what is achievable, with wide gaps between average and HEALTHY LIVES benchmark performance persisting across dimensions. The health system as a whole scores only 65 in 2008—35 O V ER V IEW percent below the benchmarks of best performance. Compared with top-performing countries and states, the Average dimension scores ranged from a low of 53 for U.S. as a whole is falling short in promoting healthy, long, efficiency to 72 for healthy lives (Exhibit 1). and productive lives for everyone. The Scorecard includes The overall score for U.S. health system performance five indicators in this dimension, including potentially failed to improve from the 2006 to the 2008 Scorecard. preventable deaths, infant mortality, disability, and healthy Access to care significantly declined due to continuing life expectancy. From 2006 to 2008, average performance erosion in health insurance coverage and affordability. declined from 75 to 72, due to poor performance on two Across the 37 core indicators, performance scores more core indicators. The score reflects the growing gaps in health often worsened than improved, primarily because outcomes between average and top performance, particularly of declines in national rates. Among the 37 indicator as the U.S. lags behind gains achieved by leading countries. scores, 41 percent of scores declined, about a third (35%) Appendix A Table 2 presents the national rate, range of improved, and the rest exhibited no change (or were not performance, and scores for indicators in this dimension. updated). Looking at underlying national averages for all indicators, nearly half showed little or no change, and P RE V EN T A B LE MOR T ALI T Y about as many declined as improved between the 2006 The U.S. fell into last place among 19 industrialized and 2008 Scorecard (see Appendix A Table 1).13 countries on national rates of mortality considered Performance remains uneven within U.S. borders— “amenable to health care.”14 These are deaths before age with up to fivefold variation (twofold variation on average) 75 caused by at least partially preventable or treatable between the top- and bottom-tier states, health care conditions, such as bacterial infections, screenable facilities, or health plans (see Appendix A). Moreover, the cancers, diabetes, heart disease, stroke, and complications range of performance within the U.S. more often widened of common surgical procedures. While the U.S. rate than narrowed from the 2006 to 2008 Scorecard. Equity improved 4 percent between 1997–1998 and 2002–2003 gaps also persisted between advantaged and disadvantaged (from 115 to 110 deaths per 100,000), rates improved by groups. On key indicators, the bottom of the performance 16 percent on average in the other countries (Exhibit 4). range would have to improve by 40 percent on average In fact, countries that began with considerably higher simply to reach current national rates of performance, premature mortality rates than the U.S., including the which are often only mediocre. United Kingdom, Ireland, and Portugal, now outperform Although there are encouraging pockets of the U.S. in preventing or delaying such deaths. At the same improvement, the U.S. still has a long way to go to make time, the top three countries (France, Japan, and Australia) its health system the best possible. The country as a whole have raised the bar of performance. As a result, U.S. death is often failing to keep pace with levels of performance rates are now 59 percent higher than in countries with the attained by leading nations, states, and delivery systems, lowest rates. Improving U.S. mortality from amenable 18 H E A LT H Y L I V E S EXHIBIT 4 Mortality Amenable to Health Care Deaths per 100,000 population* 1997/98 2002/03 150 134 130 128 116 115 113 115 109 110 106 104 99  101 103 103 100 97  97 93 96 88 89  89  88 90 81 84  82 82 84 84 77 80  76 74  74  71 71 65 50 0 ce n ia n ly da ay s n ce r ia y d d k m nd l s ga te nd an ar pa de ai an an l I ta do an ee rw ra na la st a tu Sp nm rm la Ja nl al e St st I re Au ng Fr Gr No Ca Sw r er Ze Fi Po Au Ge De d Ki th i te w Ne d Ne Un i te Un * Countries’ age-standardized death rates before age 75; includes ischemic heart disease, diabetes, stroke, and bacterial infections. See Appendix B for list of all conditions considered amenable to health care in the analysis. Data: E. Nolte and C. M. McKee, London School of Hygiene and Tropical Medicine analysis of World Health Organization (WHO) mortality files (Nolte and McKee 2008). Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 causes to levels achieved by these leading countries would advanced countries, even as it reached a new high of almost translate into 101,000 fewer deaths per year. 78 years in 2006.17 The U.S. ranks poorly in terms of healthy The rate of infants born in the U.S. who die before life expectancy at age 60, as U.S. adults spend more of their their first birthday improved slightly from 2002 to 2004 lives in poor health than adults in other countries. Perhaps (from 7.0 to 6.8 deaths per 1,000 live births), thus returning this is not surprising, given the greater burden of chronic to earlier levels. Yet, the U.S. average remains well above health problems among older adults in the U.S., compared rates in the lowest states and countries. Rates of infant with adults abroad, and the adverse health consequences mortality in the worst-performing states are more than for older adults after long periods without insurance.18 twice those in benchmark states. Of concern, the gap Activity limitations. More than one of every six between the leading and lagging states grew wider in working-age adults (18%) reported being unable to work 2004, as states with the highest rates—primarily poor and or carry out everyday activities because of health problems located in the South—experienced an increase in infant in 2006—up from 15 percent reporting limitations in 2004. mortality.15 Moreover, the U.S. ranked last among eight Health-related limitations increased in both the top and industrialized countries that report infant mortality using bottom five states, but the deterioration was greatest in the same methodology, with a national rate more than the bottom states. Previously reported rates at which double the leading countries (2.8 to 3.1 deaths per 1,000 children miss large numbers of school days because of live births in Japan, Iceland, and Sweden in 2004).16 illnesses or injuries vary more than twofold across states. These findings indicate the need for better prevention IM P A C T S O F P OOR HEAL T H and treatment of chronic diseases to enhance quality of Healthy life expectancy. Reflecting these mortality trends, life and capacity to work, particularly among younger life expectancy in the U.S. has not kept pace with other cohorts as they age. 19 Indeed, there is much room for improvement on performance.19 Appendix A Tables 3 and 4 present the the nation’s ability to promote health and well-being and national rate, range of performance, and scores for each much to gain from cultivating a healthy and productive indicator in this dimension. workforce. The U.S. is unlikely to move forward on this central goal unless the health system’s shortcomings in E F F E C T I V E C ARE terms of health care access, quality, and efficiency are Across five indicators measuring whether Americans addressed simultaneously. receive services that are effective and appropriate for preventing or treating a given condition and controlling QUALIT Y OF CARE chronic illness, the average score increased from 74 to 78. Two indicators showed substantial progress in narrowing O V ER V IEW the gap between average and benchmark performance, The nation is not making consistent progress in improving while three exhibited no or little improvement. the quality of health care based on Scorecard indicators Preventive care. Only half of adults received all age- that track the extent to which patients receive care that appropriate preventive care such as immunizations, cancer is effective, safe, well-coordinated, timely, and patient- screenings, and blood pressure and cholesterol tests in 2005, centered. Although national average performance improved with no improvement since 2002 (Exhibit 5). Achieving the for several indicators of effective care, in particular control Scorecard’s benchmark target of 80 percent would mean of chronic disease and care in the hospital, there was no that 70 million more adults would receive all recommended improvement in overall receipt of adult preventive care. preventive care. Multifaceted interventions, including Performance was uneven, slipped, or did not keep pace organizational changes, are needed to make delivery of with benchmarks of safe, well-coordinated, and patient- preventive services a routine part of patient care.20 centered care. As a result, the average of these key areas The proportion of young children who received all comprising quality failed to improve, yielding an overall recommended doses of five key vaccines increased slightly, score of 71—an average of 29 percent below benchmark from 79 percent in 2003 to 81 percent in 2006, although Q UA LIT Y: EFFEC TI V E C A RE EXHIBIT 5 Receipt of Recommended Screening and Preventive Care for Adults Percent of adults (ages 18+) who received all recommended screening and preventive care within a specific time frame given their age and sex* U.S. Average 2002 49 2005 50 U.S. Variation 2005 400%+ of poverty 58 200%-399% of poverty 47 <200% of poverty 39 Insured all year 53 Uninsured part year 46 Uninsured all year 32 0 20 40 60 80 100 * Recommended care includes seven key screening and preventive services: blood pressure, cholesterol, Pap, mammogram, fecal occult blood test or sigmoidoscopy/colonoscopy, and flu shot. See Appendix B for complete description. Data: B. Mahato, Columbia University analysis of Medical Expenditure Panel Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 20 Q UA LIT Y: EFFEC TI V E C A RE EXHIBIT 6 Chronic Disease Under Control: Diabetes and Hypertension Diabetes Hypertension Percent of adults (ages 18+) with diagnosed Percent of adults (ages 18+) with hypertension diabetes whose HbA1c level <9.0% whose blood pressure <140/90 mmHg 100 100 88 79 75 75 50 50 41 31 25 25 0 0 1999–2000 2003–2004 1999–2000 2003–2004 Data: J. M. McWilliams, Harvard Medical School analysis of National Health and Nutrition Examination Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 the rate declined among top-performing states. This area control of blood sugar (hemoglobin A1c less than of relatively better performance demonstrates the value of 9%) increased from 79 percent to 88 percent from population health improvement policies, such as school 1999–2000 to 2003–2004. Many diabetics need to vaccination requirements, coupled with a commitment lower their blood sugar levels further to achieve good to measuring and improving rates at national, state, and control (hemoglobin A1c less than 7%), which just community levels. more than half (56%) have achieved. Mental health care. Among adults who had major •• Control of high blood pressure increased from 31 depressive episodes, rates of those receiving at least some percent to 41 percent of adults with hypertension—a treatment increased marginally, from 65 to 69 percent risk factor for heart disease and stroke—during the from 2004 to 2006, leaving nearly one-third without same time period. any care. (A measure of mental health care for children Further improvement could help prevent or delay could not be updated.) Further pointing to gaps in mental serious disease complications. Achieving the level of health care, research finds that mental health treatment is control seen in the best-performing health plans could often inadequate, even among those who do receive it.21 prevent up to 39,000 premature deaths and save up to Improving depression care would not only improve quality $2 billion in medical costs annually.23 of life for individuals, it would also increase workplace Rates of control of these two common conditions productivity by an estimated $2.2 billion annually.22 vary widely across health plans, with a 30 to 60 percentage Chronic disease management. According to the results point spread between top- and bottom-performing plans. of physical exams conducted on a nationally representative Moreover, national rates of control vary significantly sample, rates of control of two common chronic conditions, depending on whether adults have insurance. Uncontrolled diabetes and hypertension, have improved (Exhibit 6). diabetes rates (HbA1c 9% or higher) were 37 percent •• Among adults with diabetes, rates of at least fair among the uninsured, compared with 19 percent among 21 insured diabetics during 1999–2004. Among adults with Web site.24 This initiative changed the landscape for hypertension, 79 percent of uninsured adults had blood hospital acceptance and reporting of quality performance pressure levels that were not under control, compared with and sparked broad efforts to improve. Top hospitals are 59 percent of the insured (see Scorecard Chartpack). achieving 100 percent on these basic process measures, Hospital care for common conditions. Hospitals indicating that full adherence to guidelines is possible. delivered 10 evidence-based treatments 90 percent Researchers estimate that if hospitals in the bottom quartile of the time to patients with heart attack, heart failure, of performance improved to the level of the top quartile, and pneumonia in 2006—up from 84 percent in 2004 more than 2,000 deaths could be avoided each year.25 (Exhibit 7). Although the entire distribution moved up, the spread between the bottom and top 10th percentiles C OORDINA T ED C ARE of hospitals remained wide, particularly for pneumonia Poor care coordination continues to be pervasive in the and heart failure, for which there were gaps of 20 to 30 U.S., owing to a fragmented delivery system and lack of percentage points, respectively, between leading and incentives for integration. The average score across five lagging hospitals. indicators of care coordination slipped from 72 to 71, with The positive general trend on hospital quality indicators only one indicator improving. Better coordination of reflects the influence of national consensus on a single set patient care throughout the course of treatment and across of measures, widespread hospital data reporting following sites of care would help ensure appropriate treatment linkage to Medicare payment updates, and public reporting and follow-up, minimize the risk of error, and prevent of hospital-specific results on the federal Hospital Compare complications leading to costly emergency department Q UA LIT Y: EFFEC TI V E C A RE EXHIBIT 7 Hospitals: Quality of Care for Heart Attack, Heart Failure, and Pneumonia Overall Composite for All Three Conditions Individual Composites by Condition, 2006 Percent of patients who received recommended Percent of patients who received recommended care for all three conditions* care for each condition* 2004 2006 Median 90th percentile 10th percentile 99 100 99 98 100 96 100 96 95 90 91 91 88 87 84 78 76 75 75 75 71 50 50 25 25 0 0 Median Best 90th 10th Heart attack Heart failure Pneumonia percentile percentile * Composite for heart attack care consists of 5 indicators; heart failure care, 2 indicators; and pneumonia care, 3 indicators. Overall composite consists of all 10 clinical indicators. See Appendix B for description of clinical indicators. Data: A. Jha and A. Epstein, Harvard School of Public Health analysis of data from CMS Hospital Compare. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 22 Q U A L I T Y : C O O R D I N AT E D C A R E EXHIBIT 8 Transition Care: Hospital Discharge and Follow-Up Care for Chronically Ill Patients Percent of heart failure patients Percent of patients hospitalized for mental illness discharged home with written instructions* with follow-up within 30 days after discharge, 2006 2004 2006 Mean 90th percentile 10th percentile 100 100 94 87 88 81 80 76 75 75 68 63 56 58 50 50 50 36 29 25 25 17 9 0 0 U.S. mean 90th percentile 10th percentile Private Medicare Medicaid Hospitals Managed Care Plans * Discharge instructions must address all of the following: activity level, diet, discharge medications, follow-up appointment, weight monitoring, and what to do if symptoms worsen. Data: Heart failure discharge instructions—A. Jha and A. Epstein, Harvard School of Public Health analysis of data from CMS Hospital Compare; follow-up after hospitalization for mental illness—Healthcare Effectiveness Data and Information Set (NCQA 2007). Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 visits and hospital admissions. There are additional nation of care at the time of hospital discharge helps benefits to patients, including reduced stress and confusion prevent subsequent complications and readmissions, surrounding their treatment and time saved in navigating especially for patients with complex or chronic conditions.28 a complex health system. Proper hospital discharge planning ensures that patients Regular source of primary care. Connection to a source understand what to do when they get home and whom of primary care can facilitate care coordination as well as to call if they have questions or concerns, and facilitates provide preventive care and chronic care management. Yet arrangements for follow-up care. In 2006, two-thirds (68%) in 2005, more than one-third (35%) of nonelderly adults of patients hospitalized with heart failure received complete reported they did not have an easily accessible primary care written instructions at discharge, a significant increase from provider that acts as a central source of care and referrals; the 50 percent in 2004 (Exhibit 8). Yet, one-third, on average, rate was the same in 2002.26 Those who lack a usual source still left without discharge instructions. Although variation of primary care are more likely to have unmet health care narrowed across hospitals, a nearly threefold difference needs, to be hospitalized, and to have higher costs of care remained from the top to bottom groups. Disturbingly, and are less likely to keep doctors’ appointments, adhere only one-third of patients in the worst-performing hospitals to treatment, and receive preventive care.27 Having health received full discharge instructions. insurance is a key factor for ensuring access to primary care: Follow-up after a hospitalization for mental illness individuals who are insured all year have a primary care supports a patient’s transition to the community and connection at twice the rate of those who are uninsured. can help avoid further acute crises.29 Yet, such follow- Coordination of care for hospital patients. Coordi­ up failed to occur in one of every four cases in private 23 Q U A L I T Y : C O O R D I N AT E D C A R E EXHIBIT 9 Nursing Homes: Hospital Admission and Readmission Rates Among Nursing Home Residents Percent of long-stay residents Percent of short-stay residents re-hospitalized with a hospital admission within 30 days of hospital discharge to nursing home 40 2000 2004 40 2000 2004 27 26 23 21 22 22 19 19 20 20 20 18 17 17 16 15 15 13 14 13 12 11 0 0 Median 10th 25th 75th 90th Median 10th 25th 75th 90th percentile percentile percentile percentile percentile percentile percentile percentile Data: V. Mor, Brown University analysis of Medicare enrollment data and Part A claims data for all Medicare beneficiaries who entered a nursing home and had a Minimum Data Set assessment during 2000 and 2004. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 health plans and in two of every five cases in Medicare and above benchmark rates. There is more than a twofold Medicaid health plans in 2006 (Exhibit 8). Moreover, the difference in performance between the lowest and highest rate declined by 8 percent among Medicare health plans quartile of agencies (19% to 48%). from 2004 to 2006, widening the performance deficit High rates of unnecessary hospitalizations put frail with private plans. Rates of follow-up care among patients elders at risk of poor outcomes or complications that often with mental illness varied fivefold between the best- and lead to subsequent deteriorations in their conditions. worst-performing health plans, and average rates have As discussed in recent Medicare Payment Advisory failed to improve over time. Commission reports, the high rates of potentially avoidable Hospitalization of nursing home residents and home readmissions signal a need to focus on improving the health patients. Nursing homes and home health agencies quality of nursing care, discharge planning, and transition can limit hospitalization rates by working with hospitals care.31 Yet, current payment incentives often work against and physicians to coordinate care and by providing high- these goals. In the case of those eligible for Medicare quality care to avoid complications that require acute and Medicaid, conflicting payer incentives leave neither care.30 Trends are moving in the wrong direction. Almost program accountable or with incentives to manage care one of five long-term nursing home residents (19%) were well for frail elderly or disabled residents.32 hospitalized in 2004, up from 17 percent in 2000 (Exhibit 9). Likewise, 18 percent of hospitalized patients who were SA F E C ARE discharged to a nursing facility were readmitted to the Patient safety risks remain high. National rates across hospital within 30 days in 2004, up from 17 percent in 2000. several safety indicators improved, yet they did not Rates increased in both low- and high-rate states. Among keep pace with gains made by benchmark performers. home health care patients, the national hospitalization Therefore, the average score among five safety indicators is rate remained at 28 percent from 2004 to 2006–2007, well only 68 out of 100. Nearly one-third (32%) of U.S. patients 24 surveyed in 2007 said that, in the last two years, a medical by evidence-based guidelines to determine if they had a mistake or a medication or lab test error was made during bacterial infection warranting antibiotic treatment. This their care, with little change from 2005. It would take a 40 rate improved from 43 percent during the years 1997 percent reduction to reach the low level of errors reported to 2003. Variation among health plans reporting to the in the benchmark country (Germany). National Committee for Quality Assurance remains wide, Drug safety. Drug safety is of particular concern. especially among Medicaid plans. Among patients living in the community, the rate of Nursing home pressure sores. One of eight high-risk adverse drug effects serious enough to require a visit nursing home residents and one of six short-stay residents to the doctor or a hospital emergency department develop pressure sores, which suggests they are receiving increased by one-third from 2001 to 2004. Regional inadequate care. Pressure sores carry the risk of serious variation widened. Patient injuries may be caused by complications, including death. The improvements side effects of the drugs or from human and system achieved through collaborative initiatives and in individual failures, such as inadequate patient education, facilities suggest that it is possible to substantially reduce inadequate monitoring of high-risk drugs, and gaps in the incidence of pressure sores.33 Yet, average rates showed coordination of care. no or little improvement from 2004 to 2006 and remained In 2004, nearly one of five elderly Americans (17%) highly variable across states. It would take a 34 percent was prescribed one of the 33 drugs that experts consider reduction in national pressure sore rates to reach the level potentially inappropriate for the elderly because of limited achieved in the top 10 percent of states. effectiveness or risk of harm. There was little change in the Hospital mortality. The hospital standardized mortality national rate since 2002. ratio (HSMR) is the only safety indicator included in the Overuse of antibiotics puts all patients at risk from the Scorecard for which there has been broad improvement. threat of antibiotic-resistant pathogens. In 2004, more than Based on Medicare data, this risk-adjusted mortality one-third (35%) of children prescribed an antibiotic for ratio declined 19 percent, from 101 in 2000–2002 to 82 in a sore throat did not receive a “strep” test recommended 2004–2006 (Exhibit 10). The HSMR is a ratio of actual Q UA LIT Y: SA FE C A RE EXHIBIT 10 Hospital-Standardized Mortality Ratios Standardized ratios compare actual to expected deaths, risk-adjusted for patient mix and community factors.* Medicare national average for 2000=100 140 Ratio of actual to expected deaths in each decile (x 100) 2000–2002 2004–2006 120 118 112 103 106 106 101 100 100 94 97 93 86 89 82 85 83 83 85 78 78 79 81 80 74 60 40 20 0 U.S. Mean 1 2 3 4 5 6 7 8 9 10 Decile of hospitals ranked by actual to expected deaths ratios * See Appendix B for methodology. Data: B. Jarman analysis of Medicare discharges from 2000 to 2002 and from 2004 to 2006 for conditions leading to 80 percent of all hospital deaths. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 25 deaths to expected deaths; expected death rates are decisions, and improve outcomes of care.39 The overall generated based on average national mortality in 2000, score for patient-centered and timely care declined with adjustments made for patient and community risk from 72 to 69, as two indicators declined (one could not factors.34 Acceleration in the decline of hospital risk- be updated). The Scorecard results indicate that there adjusted mortality was noted first in 2002 and was sustained are major deficiencies in providing timely care and throughout the reporting period. This improvement communicating effectively with patients. National scores followed widespread availability of risk-adjusted measures on indicators are as much as 65 percent below benchmarks and several high-profile local and national initiatives that set by leading countries, health plans, or hospitals. aimed to improve hospital patient safety and reduce Rapid access to primary care. U.S. adults with health mortality by focusing on actionable strategies to track and problems are significantly less likely than patients in five of improve hospital quality. High-profile national initiatives the seven countries surveyed to get a rapid appointment include the Joint Commission National Patient Safety with a physician—the same or the next day—when they Goals, The Institute for Healthcare Improvement’s 100,000 are sick. Only 46 percent of patients reported having such Lives and 5 Million Lives Campaigns,35 The Leapfrog rapid access in 2007, nearly the same as in 2005. The Group’s Hospital Quality and Safety Survey, the National failure to improve highlights the slow pace of adoption Surgical Quality Improvement Program, 36 and the Center of advanced access models of care in physician practices for Disease Control and Prevention’s National Healthcare and clinics. The U.S. rate would need to improve by more Safety Network, and others. than 75 percent to reach the benchmark rate (81%). Eight years after the Institute of Medicine issued After-hours care. U.S. adults with health problems its national call to action on patient safety, the federal are also the most likely among adults in seven countries government is finally moving to establish Patient Safety surveyed to report difficulty obtaining health care after Organizations with the capacity to collect, analyze, and hours without going to the emergency department. This report on safety events at the national level. In a recent rate increased from 61 percent to 73 percent from 2005 to survey, physicians said they are willing to share their 2007 (Exhibit 11). Studies in the U.S. indicate that improved experiences with medical errors for learning purposes, after-hours care and better access to primary care can reduce but they find the current error reporting systems the need for relatively costly emergency department visits, inadequate. Studies conclude that health care institutions particularly among higher-risk, low-income patients.40 need to do more to engage physicians in meaningful Physician communication. Open and clear reporting leading to demonstrable improvement at the communication between doctors and their patients is a key local level.37 component of patient-centered care. On average, just over Collaborative efforts to bring down infection rates half of U.S. patients in 2004 and 2002 (57% and 54%) said in intensive care units have shown that following simple their doctors always listened carefully, explained things “checklists” or “bundles” of evidence-based practices clearly, showed them respect, and spent enough time with can reduce rates to zero—setting new benchmarks for them. Patient communication experiences vary widely by performance.38 This level of perfection is being achieved insurance status and source of coverage. The national rate by the top 10 percent to 25 percent of intensive care in 2004 remained well below the 75 percent benchmark units participating in the Centers for Disease Control rate set by top-performing health plans. Interventions and Prevention’s National Healthcare Safety Network, aimed at both physicians and patients may improve the a federal benchmarking initiative (see Scorecard quality of interpersonal medical interactions.41 Chartpack). Wider adoption of these initiatives, coupled Hospital responsiveness to patients. A wide range with Medicare’s new policy of refusing to pay for certain in performance persisted among hospitals on three preventable errors, may accelerate improvements in indicators of patient-centered hospital care, with a 15-to- hospital patient safety. 24-percentage-point difference between the top decile of hospitals (rates of 66% to 75%) and the bottom decile (rates P A T IEN T - C EN T ERED AND T IMELY C ARE of 48% to 60%) on measures of how well staff manage Patient-centered care and timely access to care can increase pain, respond when patients press a call button or need adherence to treatment plans, help engage patients in care help going to the bathroom, or explain medications and 26 Q U A L I T Y : P A T I E N T- C E N T E R E D , T I M E L Y C A R E EXHIBIT 11 Difficulty Getting Care on Nights, Weekends, Holidays Without Going to the Emergency Room, Among Sicker Adults Percent of adults who sought care reporting “very” or “somewhat” difficult 100 2005 2007 73 75 68 69 61 61 48 49 50 50 25 0 United States Netherlands New Zealand Germany United Kingdom Australia Canada International Comparison Data: 2005 and 2007 Commonwealth Fund International Health Policy Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 Q U A L I T Y : P A T I E N T- C E N T E R E D , T I M E L Y C A R E EXHIBIT 12 Patient-Centered Hospital Care: Staff Managed Pain, Responded When Needed Help, and Explained Medicines, by Hospitals, 2007 Percent of patients reporting “always” Mean Best hospital 90th percentile hospitals 10th percentile hospitals 100 97 96 91 75 75 72 67 66 60 60 58 48 49 50 25 0 Staff managed pain well* Staff responded when needed help** Staff explained medicines and side effects*** * Patient’s pain was well controlled and hospital staff did everything to help with pain. ** Patient got help as soon as wanted after patient pressed call button and in getting to the bathroom/using bedpan. *** Hospital staff told patient what medicine was for and described possible side effects in a way that patient could understand. Data: CAHPS Hospital Survey (Retrieved from CMS Hospital Compare database at www.hospitalcompare.hhs.gov). Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 27 their possible side effects (Exhibit 12). The best hospitals departments for primary care, duplication of services, achieved very high rates of patients giving top ratings on and failure to follow-up on test results or preventive care. these questions, illustrating that it is possible for hospitals Rising numbers of uninsured as well as escalating health to do much better in meeting patients’ needs. care costs and health insurance premiums create barriers These results from 2,500 hospitals participating in the to care and place financial strain on insured as well as Hospital Consumer Assessment of Healthcare Providers uninsured patients.42 Reflecting these trends, performance and Systems (HCAHPS) Survey are remarkably similar on four of five access indicators declined substantially, to findings from a smaller pilot of the survey reported as increasing numbers of middle- as well as low-income in the first National Scorecard. The public release of families found themselves at risk of inadequate access to these data on the Medicare Web site in March 2008 care. The overall score on this dimension dropped from marks a turning point—the first time that consumers 67 to 58—further from the goal of full participation and have been able to compare hospital performance on a affordable access.43 Appendix A Table 5 presents the uniform patient survey. It also shows the positive role national rate, range of performance, and scores for each government can play in promoting greater accountability indicator in this dimension. by sponsoring the development of a standard survey and influencing providers to participate through Medicare P AR T I C I P A T ION payment incentives. To date, most of the erosion in insurance coverage has occurred among working-age adults. Based on annual HEALTH CARE ACCESS census data, from 1999–2000 to 2005–2006 the number of states where 23 percent or more of the working-age adult O V ER V IEW population is uninsured grew from two to nine, while Access to care is fundamental to high-quality care. the number of states with less than 14 percent uninsured Inadequate access can result in inefficient care from declined from 22 to eight (Exhibit 13). Children fared avoidable complications, reliance on emergency better due to public coverage expansions. In 2005–2006, A C C E S S : P A RT I C I P AT I O N EXHIBIT 13 Percent of Adults Ages 18–64 Uninsured by State 1999–2000 2005–2006 WA WA ME ME MT ND MT ND MN VT MN VT OR OR NH NH ID WI NY MA ID WI NY MA SD SD MI CT RI MI CT RI WY WY NJ NJ IA PA IA PA NE OH NE OH NV MD NV MD IL IN DE IL IN DE UT UT WV WV CO CO CA KS MO VA DC CA KS MO VA DC KY KY NC NC TN TN AZ OK AZ OK NM AR SC NM AR SC MS AL GA MS AL GA LA LA TX TX FL FL AK AK 23% or more 23% or more HI 19%–22.9% HI 19%–22.9% 14%–18.9% 14%–18.9% Less than 14% Less than 14% Data: Two-year averages 1999–2000, updated with 2007 Current Population Survey correction, and 2005–2006 from the Census Bureau’s March 2000, 2001 and 2006, 2007 CPS. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 28 A C C E S S : P A RT I C I P AT I O N EXHIBIT 14 Uninsured and Underinsured Adults, 2007 Compared with 2003 Percent of adults (ages 19–64) who are uninsured or underinsured 100 Underinsured* Uninsured during year 72 75 68 19 24 50 42 35 14 27 9 25 49 48 17 11 26 28 4 13 16 0 2003 2007 2003 2007 2003 2007 Total Under 200% of poverty 200% of poverty or more * Underinsured defined as insured all year but experienced one of the following: medical expenses equaled 10% or more of income; medical expenses equaled 5% or more of income if low income (<200% of poverty); or deductibles equaled 5% or more of income. Data: 2003 and 2007 Commonwealth Fund Biennial Health Insurance Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 only five states had more than 16 percent of children because of costs. In contrast, only 5 percent of adults in uninsured, down from nine in 1999–2000. And in twelve the Netherlands, the benchmark country, reported such states, fewer than 7% of children were uninsured.44 financial barriers to care. The Netherlands has universal As the number of adults without insurance has steadily coverage with a broad range of benefits and modest cost- grown, so has the number of “underinsured”—those who sharing by U.S. standards. are insured all year but have medical bills or deductibles that were high relative to their incomes.45 In 2007, 25 A F F ORDA B LE C ARE million adults (14%) were underinsured, an increase of The costs of both health insurance and medical care have more than 60 percent since 2003 when 16 million were become less affordable. The average cost of family coverage underinsured. This sharp jump was driven by a near obtained at employer group rates exceeded $12,000 a year tripling in the rate (from 4% to 11%) among those with in 2007.46 With premiums rising faster than wages, the moderate or higher incomes (200% of the federal poverty average cost of insurance premiums relative to income level or more). Another 50 million adults were uninsured increased in almost all states. As a result, the percent during the year. As a result, as of 2007, more than 75 million of adults residing in a state where employer premiums adults—42 percent of all adults ages 19 to 64—were either averaged less than 15 percent of the median household uninsured during the year or underinsured, up from 35 income declined precipitously, from 58 percent to 25 percent in 2003 (Exhibit 14). percent over the most recent two years. Although low-income adults remain most at risk, By 2005, nearly one of four adults under age 65 the increase in the percent uninsured or underinsured (23%) lived in families with high out-of-pocket health was greatest among those with incomes of 200 percent care costs, including premiums and direct spending for of poverty or higher. services, up from 19 percent in 2001. This increase was The erosion in coverage undermines access to care. driven entirely by rising costs among those with private In 2007, more than one-third of U.S. adults (37%) went insurance. Financial burdens were especially steep among without needed care, including prescription drugs, people who purchased insurance in the nongroup market: 29 ACCESS: AFFORDABLE C ARE EXHIBIT 15 Medical Bill Problems or Medical Debt Percent of adults (ages 19–64) with any medical bill problem or outstanding debt* National Average By Income and Insurance Status, 2007 100 100 Insured all year Uninsured during year 75 75 68 61 56 41 50 50 45 34 33 29 25 25 0 0 2005 2007 Total Under 200% 200% of poverty of poverty or more * Problems paying or unable to pay medical bills, contacted by a collection agency for medical bills, had to change way of life to pay bills, or has medical debt being paid off over time. Data: 2005 and 2007 Commonwealth Fund Biennial Health Insurance Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 half faced high out-of-pocket burdens, compared with 40 opportunity to receive high-quality care.47 Uninsured percent in 2001. people often fail to get timely and appropriate care when Efforts to moderate premium growth have led to limits needed, leading to worse health outcomes and more costly on benefits and higher cost-sharing. The resulting exposure emergency or acute care later on. When they do get care, to costs added to the share of families who struggle with the uninsured also experience more medical errors or medical debt and medical bills. By 2007, two of five U.S. coordination problems, such as delays in transferring adults (41%) reported having problems paying medical bills, medical records/test results and duplication of tests. A being contacted by collection agencies, or paying medical recent study estimates the death toll from being uninsured debt over time, up from 34 percent in 2005 (Exhibit 15). amounted to 137,000 from 2000 to 2006, including 22,000 Having insurance is no longer a guarantee of financial deaths in 2006.48 protection: one of three (33%) adults ages 19 to 64 who were Studies also find that high uninsured rates continually insured faced medical bill problems; middle- undermine the quality of care for entire communities and lower-income adults were the most at risk. and states.49 States and communities in which large shares of the population are uninsured exhibit lower A C C ESS AND I T S RELA T IONSHI P quality and worse patient care experiences across a T O Q UALI T Y AND E F F I C IEN C Y range of care settings for insured as well as uninsured Reduced access to care has serious implications for patients, compared with communities with low rates overall health system performance. Without adequate of uninsured residents. The connection between worse coverage and financial protection, there is diminished access and lower quality is likely due to spillover 30 effects and lack of policies and practices that focus INA P P RO P RIA T E , WAS T E F UL , on community-wide population health and quality OR F RAGMEN T ED C ARE of care. Ensuring universal access to care can provide In the U.S., payment incentives can encourage physicians a foundation to improve quality and achieve more and hospitals to “do more,” even though this may mean efficient care over time.50 that patients receive services of marginal or no value.51 An example of this is the use of imaging tests for lower back pain EFFICIENCY OF THE HEALTH SYSTEM within 28 days of onset, when the patient has no apparent risk factors or sign of serious pathology. Within managed O V ER V IEW care plans, average rates for this indicator of potentially An efficient care system seeks to maximize health outcomes inappropriate testing are 50 percent higher than rates and quality for the resources spent and to enhance value achieved by benchmark health plans, with little or no change over time. Lack of access, poorly coordinated or fragmented from 2004 to 2006. Health plans have recently been stepping care, and ineffective care add cost and decrease value. up efforts to review and reduce the use of advanced imaging They also waste patients’ time. Comparisons with other services in response to their rapid proliferation.52 countries as well as regional variations in cost and quality In a cross-national survey, 22 percent of U.S. adults within the U.S. across an array of efficiency indicators all with health problems reported that test results and medical point to opportunities to achieve savings and/or improve records were not available at the time of their medical value. Overall, performance on indicators of efficiency appointment in 2007, compared with the benchmark of 9 remains especially low, with the U.S. average score of just percent in the Netherlands (Exhibit 16). U.S. patients were 53 basically unchanged from 52 in the 2006 Scorecard. The five times more likely to say that doctors unnecessarily failure to improve makes efficiency the dimension with repeated tests, as compared with patients in the benchmark the greatest gap between U.S. performance and achievable country (20% in the U.S. vs. 4% in the Netherlands). benchmarks. Appendix A Table 6 presents the national There was little change in these indicators from 2005 rate, range of performance, and scores for each indicator to 2007. Better performance in the benchmark country in this dimension. likely reflects more integrated care and widespread use of electronic medical records. EFFICIENCY EXHIBIT 16 Test Results or Medical Records Not Available at Time of Appointment, Among Sicker Adults Percent reporting test results/records not available at time of appointment in past two years 30 2005 2007 23 22 20 18 17 17 14 12 10 9 0 United States Netherlands Germany New Zealand Australia United Kingdom Canada International Comparison Data: 2005 and 2007 Commonwealth Fund International Health Policy Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 31 P O T EN T IALLY A V OIDA B LE HOS P I T AL USE levels would save $4 billion annually; savings would be Having timely access to primary care, during regular office commensurate in the under-65 population. hours or after hours, can avert the need for expensive visits Nearly one of five Medicare patients (18%) initially to the emergency department (ED) or admission to the hospitalized with one of a set of selected conditions was hospital and lessen the risk of medical complications. readmitted to the hospital within 30 days; there was One of five U.S. adults (21%) reported they went to the no change in this rate from 2003 to 2005 (Exhibit 17).53 ED for a condition that could have been treated by their Medicare 30-day readmission rates vary widely across regular doctor, more than three times the rate in the hospital referral regions: rates in the highest-rate regions benchmark country, Germany, where only 6 percent of are 50 percent higher than in the lowest-rate regions. patients reported such unnecessary ED use. Good care provided during a hospital stay and Ready access to high-quality, well-coordinated appropriate discharge planning, follow-up, and post-acute primary care can prevent complications and care can help prevent patients from being readmitted to the hospitalizations. Rates of potentially preventable hospital, thus reducing the total costs of care.54 A Medicare hospitalizations for ambulatory care–sensitive (ACS) Payment Advisory Commission analysis indicates that conditions vary by a multiple of two to four across states up to three-quarters of readmissions may be preventable and hospital referral regions, with associated variations with better primary care, transition care, and reduced in costs. The national hospital admission rate for heart complications from care received while hospitalized—a failure decreased 4 percent from 2002 to 2004, while the potential savings of $12 billion a year for Medicare.55 rate for pediatric asthma decreased 13 percent from 2003 to 2004; however, diabetes-related admissions remained V ARIA T ION IN Q UALI T Y AND C OS T S unchanged. Among Medicare beneficiaries, a composite In the Medicare program, the costs of care are highly rate of hospital admissions for 11 ACS conditions concentrated among patients with multiple chronic decreased 9 percent from 2003 to 2005, with decreases conditions, and such costs are increasing.56 In 2005, in both the top and bottom of the distribution. Further annual costs of care to Medicare averaged $38,000 for reducing Medicare ACS admissions to benchmark patients who had all three of the following conditions: EFFICIENCY EXHIBIT 17 Medicare Hospital 30-Day Readmission Rates Percent of Medicare beneficiaries admitted for one of 31 select conditions who are readmitted within 30 days following discharge* 30 21 20 19 20 20 18 18 16 16 15 14 10 0 2003 2005 10th 25th 75th 90th 10th 25th 75th 90th U.S. Mean Hospital Referral Region State Percentiles, 2005 Percentiles, 2005 * See Appendix B for list of conditions used in the analysis. Data: G. Anderson and R. Herbert, Johns Hopkins University analysis of Medicare Standard Analytical Files (SAF) 5% Inpatient Data. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 32 EFFICIENCY EXHIBIT 18 Quality and Costs of Care for Medicare Patients Hospitalized for Heart Attacks, Hip Fractures, or Colon Cancer, by Hospital Referral Regions, 2004 1.20 Median relative (1-year survival index, median=70%) resource use=$27,499 1.10 Quality of care* 1.00 0.90 0.80 $0 $5,000 $10,000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000 Relative resource use** * Indexed to risk-adjusted 1-year survival rate (median=0.70). **Risk-adjusted spending on hospital and physician services using standardized national prices. Data: E. Fisher, J. Sutherland, and D. Radley, Dartmouth Medical School analysis of data from a 20% national sample of Medicare beneficiaries. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 heart failure, diabetes, and chronic lung disease. This over the course of the year. In all, Medicare could save represents a 20 percent increase from 2001. Costs of care more than 9,000 lives and reduce annual costs by nearly $1 vary significantly across the country, with a twofold spread billion a year for these three conditions alone, if all other between the lowest and highest 10th percentiles of hospital U.S. regions could achieve the performance levels of the regions for any combination of these three conditions benchmark regions. (see Appendix A Table 6). Focusing on these patients offers opportunities to improve care outcomes and use INSURAN C E ADMINIS T RA T I V E C OS T S resources more efficiently. Private health insurance in the U.S. is characterized by Updated analysis of regional variations for hospitalized complex benefit and cost-sharing designs and high rates Medicare patients shows that some regions of the country of turnover in plan enrollment. Health plans also incur achieve better outcomes than other regions, and at lower significant marketing and underwriting costs. Administrative cost, through more efficient systems.57 Medicare data for costs have been increasing rapidly in the U.S.: from 2000 to patients hospitalized in 2004 for heart attacks, hip fracture, 2006, per capita administrative costs increased 68 percent, or colon cancer were used to rank hospital referral regions from $289 to $485 per person, versus a 47 percent increase in terms of their care outcomes and relative resource use in national health expenditures per capita.58 (Exhibit 18). Comparing the best- and worst-performing As a result, insurance administrative costs as a share 10th percentiles, one-year mortality rates on this composite of total national health expenditures are more than indicator of three conditions ranged from 27 percent to 33 three times higher in the U.S. than in countries with percent between the best- and worst-performing regions, the lowest rates (Finland, Japan, and Australia) and 30 while risk-adjusted annual costs ranged from $25,000 to percent to 70 percent higher than in three countries $30,000. Plotting mortality and costs for all regions shows where private insurance plays a substantial role (Germany, that a high proportion of those regions with the lowest Switzerland, and the Netherlands) (Exhibit 19). Reducing one-year mortality rates also had lower total resource costs U.S. insurance overhead to this mid-range through 33 EFFICIENCY EXHIBIT 19 Percentage of National Health Expenditures Spent on Insurance Administration, 2005 Net costs of health insurance administration as percent of national health expenditures 8 7.5 6.9 6 5.6 4.8 4.2 4.3 3.9 4 3.3 2.8 2.3 1.9 2 0 d na a b ia da sa nd y ce s* ia m an an nd tr pa te an al na la do us rm nl a tr rla er Ja Fr Ca St ng Fi A us Ge itz he d Ki A Sw te et d ni N te U ni U a 2004 b 1999 * Includes claims administration, underwriting, marketing, profits, and other administrative costs; based on premiums minus claims expenses for private insurance. Data: OECD Health Data 2007, Version 10/2007. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 EFFICIENCY EXHIBIT 20 Physicians’ Use of Electronic Medical Records Percent of primary care physicians using electronic medical records 98 2001 2006 100 92 89 79 75 50 42 28 23 25 17 0 United States Netherlands New United Australia Germany Canada Zealand Kingdom International Comparison Data: 2001 and 2006 Commonwealth Fund International Health Policy Survey of Physicians. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 34 greater standardization, streamlined functions, and patients across sites of care. At the current U.S. rate of more continuous coverage would save up to $51 billion dispersion, it would require more than 30 years to expand annually. This is enough to fund half the cost of providing such tools to all physicians. comprehensive coverage to all the uninsured in the U.S. Lowering rates to the benchmark countries would save EQUIT Y IN THE HEALTH SYSTEM more than $100 billion per year. O V ER V IEW IN F ORMA T ION SYS T EMS The health care system offers the potential to provide equal T O SU P P OR T E F F I C IEN T C ARE opportunities for all to lead healthy and productive lives, a Well-integrated electronic information systems have the core founding value of the United States. However, studies capacity to improve the delivery and coordination of repeatedly reveal pervasive disparities in health outcomes care, reduce medical errors, and provide a mechanism and care experiences across different racial, ethnic, and for tracking and assessing performance. Although use of socioeconomic groups within the U.S. electronic medical records (EMRs) by U.S. physicians Reducing and eliminating such disparities has long increased from 17 percent to 28 percent from 2001 to 2006, been a major national concern and is central to improving the U.S. lags well behind leading countries that have made care for the country as a whole. Yet, the Scorecard a system-wide commitment to invest in interoperable finds persistent and wide gaps on key indicators across information technology (Exhibit 20). In the United dimensions between vulnerable populations and their Kingdom, nine of 10 primary care practices have EMRs, benchmark reference groups, with no improvement since as do 98 percent of practices in the Netherlands. Further, the 2006 baseline—the average score was 71 in 2008 clinical data systems in these countries are more likely than compared with 70 in the 2006 Scorecard. As illustrated those in the U.S. to have advanced functions to provide in Exhibit 21, wide inequities persist for each vulnerable decision support and enable information to flow with group in healthy lives, access, quality, and efficiency. EQUITy EXHIBIT 21 equity: ratio scores for insurance, income, and race/ethnicity insured high income Compared with Compared with white Compared white Compared uninsured low income^ with Black with hispanic 2006 2008 2006 2008 2006 2008 2006 2008 eQuiT y aVer age sCore 65 66 61 61 76 75 77 81 (Number of indicators*) (18) (25) (26) (26) d i m e n s i o n aV e r ag e s healthy lives NA NA 54 55 77 77 97 97 Quality Effective Care 59 57 69 68 80 76 73 70 Safe Care 97 97 94 94 77 77 94 94 Patient-Centered, Timely Care 56 56 59 59 72 62 54 64 Coordinated and effi cient Care 55 58 63 60 61 73 58 72 access 57 61 30 32 86 87 82 87 * No updated data available for 4 indicators by insurance, 4 indicators by income, and 4 indicators by race/ethnicity; used baseline score from 2006. ^ Generally income compares either poor/near poor (<200% poverty) to those of incomes of 400% of poverty or higher or compares annual incomes of under $35,000 to incomes above $45,000. For mortality, income uses either census tract poverty rates or education level. NA=data not available Data: Appendix A Table 7 presents scores for all indicators in the Equity dimension. See Appendix B for data years and sources. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 35 On average, it would require a 19 percent to 25 of having untreated dental caries than whites, and here percent reduction in the risk of poor health outcomes again minority rates are on the rise, specifically among and inadequate or inefficient care for black or Hispanic the elderly (Exhibit 22). Disparities in care experiences in minorities to reach the same rates as whites. Gaps in part reflect minorities’ lower incomes and insurance gaps. performance for uninsured and low-income populations Insured, higher-income populations are generally at lower are even wider; it would require a 34 percent to 39 percent risk of poor access and care experiences. For example, rates improvement on average to achieve parity with insured of caries going untreated are more than two times higher and high-income populations, respectively. While some among the uninsured than the insured across all ages. gaps are closing, a significant proportion have worsened Inequitable access to quality health services, or stayed the same. Moreover, in some instances, gaps especially among the chronically ill, contributes to have narrowed only because experiences grew worse for disparate short- and long-term health outcomes between white, insured, or higher-income groups. whites and minorities. The proportion of diabetic blacks and Hispanics with uncontrolled blood sugar is DIS P ARI T IES IN INSURAN C E C ON T RI B U T E more than two times that of whites, and this disparity T O DIS P ARA T E C ARE E X P ERIEN C ES has grown as white rates have fallen. Moreover, blacks Overall, minorities are much less likely than whites to get and Hispanics suffer disproportionately high rates of preventive care or proper treatment when needed; for some death and hospitalizations because of diabetes-related indicators, relative disparities are widening. For instance, complications. There has been only modest improvement blacks and Hispanics are less likely to receive treatment for in utilization of appropriate diabetes services and exams. depression than whites, and these rates have worsened at For patients, successfully managing a chronic condition the same time that there has been improvement among requires an ongoing relationship with a primary care office whites. Minorities are also at significantly higher risk that can provide easy access and organized care.59 As such, EQ U IT Y: EFFEC TI V E C A RE EXHIBIT 22 Untreated Dental Caries, by Age, Race/Ethnicity, and Income, 2001–2004 Percent of persons with untreated dental caries Children ages 6–19 Adults ages 20–64 Adults ages 65–74 50 46 44 42 41 40 38 31 32 28 27 25 23 21 19 18 13 14 10 10 0 l l l ta i t e l ac k i c a n er t y er t y ta i t e l ac k i c a n er t y er t y ta i t e l ac k i c a n er t y er t y To W h B ex pov pov To W h B ex pov pov To W h B ex pov pov M M M of of of of of of + % + % + % 0 % 0 0 0 % 0 0 0 % 0 0 40 <1 40 <1 40 <1 Data: Race/ethnicity—National Health and Nutrition Examination Survey (NCHS 2007); Total and income—J. M. McWilliams, Harvard Medical School analysis of National Health and Nutrition Examination Survey. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 36 E Q U I T Y : C O O R D I N AT E D A N D E F F I C I E N T C A R E EXHIBIT 23 Ambulatory Care–Sensitive (Potentially Preventable) Hospital Admissions, by Race/Ethnicity and Patient Income Area, 2004/2005* Adjusted rate per 100,000 population Heart failure Diabetes** Pediatric asthma 1000 904 667 554 520 500 444 392 374 390 240 178 173 144 98 110 NA 0 c c c h ite a ck ni 00 + 00 0 hi te ac k ni 00 + 00 0 hi te ac k ni 00 + 00 0 Bl a 0 5, Bl a 0 5, Bl a 0 5, W isp 5, $2 W isp 5, $2 W isp 5, $2 H $4 < H $4 < H $4 < * 2004 data for diabetes and pediatric asthma; 2005 data for heart failure. **Combines 4 diabetes admission measures: uncontrolled, short-term complications, long-term complications, and lower extremity amputations. Patient Income Area=median income of patient zip code. NA=data not available. Data: Race/ethnicity—Healthcare Cost and Utilization Project, State Inpatient Databases and National Hospital Discharge Survey (AHRQ 2007); Income area—HCUP, Nationwide Inpatient Sample (AHRQ 2007, retrieved from HCUPnet at http://hcupnet.ahrq.gov). Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 disparities in the timeliness and patient-centeredness status) continue to be at significantly greater risk of death of care, which are growing between whites and blacks, from chronic diseases, including heart disease, diabetes, further exacerbate the risk of adverse outcomes for and cancer.60 Further, disparities in mortality rates are vulnerable patient populations. widening due to an increase among the least educated and Even greater gaps in care are observed by income and substantial reductions among better educated groups.61 insurance coverage than by race and ethnicity, particularly Although adequate data are not available to track and for indicators of health care access and efficiency. The poor measure health outcomes by insurance coverage, a and uninsured are at heightened risk of lacking a regular growing body of research documents a strong relationship source of primary care, experiencing medical record or test between being uninsured and mortality.62 Recent studies coordination problems, and facing medical cost burdens. have shown that, relative to the insured, uninsured cancer Difficulties with getting timely and coordinated care can patients are more likely to be diagnosed at advanced stages lead to poor health outcomes and more costly use of care. and are less likely to survive once a diagnosis is made.63 Low-income and uninsured patients are more likely than Inadequate insurance coverage is also a major those with higher incomes and insurance to go to an concern. Both the insured and uninsured are finding it emergency department for care a primary care doctor increasingly difficult to pay for their medical care; in fact, could have provided. Moreover, rates of hospitalizations more than one-third (35%) of adults who were insured for preventable conditions are two to three times higher all year went without needed care because of costs, a in low-income communities than in more affluent areas, significant increase since 2005. Moreover, the rate of and gaps are increasing (Exhibit 23). unpaid medical bills and medical debt has increased As for health outcomes, individuals with low levels among the insured and uninsured alike. Although of education (used as a proxy for low socioeconomic equity scores by income and insurance demonstrated 37 some improvements on indicators of access to care, this even as primary care doctors are expected to do more to occurred only because performance declined more rapidly promote prevention and care of chronic disease.67 As a for the high-income and insured groups. It would still result, primary care has become less attractive as a career: require about 40 percent and 70 percent improvement the proportion of residents choosing primary care has across these indicators for the disparities by insurance declined at the same time as the workforce is aging.68 and income, respectively, to be eliminated. Studies suggest that we need to rethink primary care Inequity in care is not just a social concern, but to enhance the capacity to provide accessible, quality, an issue of concern for health system performance. patient-centered, and coordinated care.69 “Medical home” Disparities undermine performance across all dimensions approaches that invest in clinical practice information of care—access, quality, and efficiency—and lead to missed systems and embed primary care in integrated systems opportunities to ensure long, healthy, and productive lives. have the potential to improve outcomes, satisfaction, and Appendix A Table 7 presents scores for each of the enable more efficient use of resources. As illustrated by vulnerable groups (i.e., uninsured, low-income, black, multiple Scorecard indicators, the U.S. needs a renewed and Hispanic populations) for indicators in the equity emphasis on primary care for the 21st century—including dimension. See methodology box on page 17 for calculation payment systems that support primary care physicians and of equity ratio scores. midlevel practitioners working together on teams—as a foundation for accessible, high-value care. SYSTEM CAPACIT Y TO An empowered nursing workforce can positively INNOVATE AND IMPROVE influence patient and nursing home resident satisfaction and quality of care.70 Job dissatisfaction among nurses contributes NO T S C ORED (see Scorecard Chartpack for data) to shortages and high staff turnover, which drive up costs The capacity to innovate and improve to achieve excellence and put patients at risk.71 Data compiled for the Scorecard is fundamental to a high-performing health care system. reveal that nursing staff satisfaction is much more variable It includes: in nursing homes than in hospitals.72 High-performing •• a care system that supports a skilled and motivated institutions offer benchmarks for improvement. health care workforce, with an emphasis on primary A growing body of evidence indicates that higher care and population health; levels of registered nurse staffing in hospitals and nursing •• a culture of quality improvement and continuous homes is associated with improved quality.73 Case studies learning that promotes and rewards recognition indicate that creative use of nurses with redesigned work of opportunities to reduce errors and improve processes can free up time to spend on patient care. Nurse outcomes; and staffing levels are of particular concern in nursing homes •• investment in public health initiatives, research, and and vary widely across states. Nursing homes in the five information necessary to inform, guide, and drive states with the highest registered nurse staffing levels health care decision-making and improvement. provide double the hours per patient day as the national On all three aspects, the U.S. currently under-invests in median, and six times more than those in the five states the system capacity to improve. with the lowest average staffing levels.74 HEAL T H C ARE WOR K F OR C E ORGANIZA T IONAL C UL T URE Countries and areas with more primary care physicians In its 1999 report, To Err Is Human, the Institute of (in proportion to population and to medical specialists) Medicine called on health care organizations to “develop achieve more equitable and better overall outcomes at a culture of safety such that an organization’s design lower cost.64 Viewed from an international perspective, processes and workforce are focused on a clear goal— the U.S. has a relatively weak primary care system.65 dramatic improvement in the reliability and safety of the Current payment mechanisms undervalue primary care process.”75 Teamwork is a key mechanism for achieving care and fail to support the time and teams necessary high reliability.76 Creating effective teamwork and a culture to manage and coordinate care.66 Notably, earnings of of safety are challenging goals for organizations to achieve. primary care physicians lag well behind those of specialists, A survey completed in 2007 by staff at 519 hospitals reveals 38 wide variation between high- and low-scoring hospitals NA T IONAL HEAL T H E X P ENDI T URES on how well they nurture teamwork and continuous F OR RESEAR C H AND P U B LI C HEAL T H improvement.77 Notably, less than half of hospitals respond Building the information infrastructure of the health care effectively to medical or medication errors. Case studies system could pay dividends in increased capacity, efficiency, find that organizational leaders can engage their workforce and quality. Necessary elements to transform care systems to promote a safety culture through persistent attention and support broader population health improvement and effort.78 include interoperable information systems, information on Instituting a culture of resident-centered care practices clinical and cost effectiveness, payment incentives aligned in nursing homes has been shown to improve residents’ with outcomes, and population health activities that can quality of life, increase staff satisfaction, and reduce staff help prevent disease and its complications. turnover.79 This approach emphasizes a “home-like” However, between 2000 and 2006, national spending environment in which residents make decisions about daily on public health activities fell behind, increasing by only activities and promotes collaborative decision-making 35 percent, as compared with a 55 percent increase in and a consistent care team made up of staff members national health expenditures and a 78 percent increase who know residents and their needs. In a 2007 survey of in insurance administrative costs. Likewise, in an era of nursing home directors of nursing, almost half expressed medical care advances, national investment in research a commitment to culture change, but only a minority regarding clinical and cost effectiveness—what works well indicated that their facility was adopting more than a for which patients and when—has failed to keep pace to few of these practices.80 Federal officials have expressed inform health care decisions. Only about 5 percent of support for culture change, but have yet to support the goal the federal research budget is devoted to health systems with changes in payment policies to enable the approach improvement research—less than $1 for every $1,000 in to become widespread.81 national health care spending. This amount is grossly out of proportion to the scope of the nation’s health system. Increased funding for comparative medical effectiveness research and improved patient decision-making would more than pay for itself, saving up to an estimated $368 billion over 10 years from more effective and efficient care.82 39 Summary and Implications Overall, the National Scorecard on U.S. Health System and coordination as well as reducing insurance overhead Performance, 2008, finds that the United States is losing costs and geographic variations in costs. ground in providing access to care and has uneven health Based on NCQA estimates, controlling diabetes care quality. The Scorecard also finds broad evidence of and blood pressure to benchmark levels could yield $1 inefficient and inequitable care. Average U.S. health system billion to $2 billion per year in savings through lower performance would have to improve by more than 50 medical costs.84 Improving depression care could increase percent on multiple indicators to reach the benchmarks. workplace productivity by an estimated $2.2 billion annually. The Medicare program could potentially save P O T EN T IAL F OR IM P RO V EMEN T : at least $12 billion a year by reducing readmissions and IM P A C T O F A C HIE V ING B EN C HMAR K S reducing hospitalizations for preventable conditions. Over The Scorecard makes a compelling case for change. Gaps $1 billion could be saved annually by providing better between average performance and benchmarks remain coordination for frail nursing home residents. large, underscoring opportunities to save lives, improve Further savings are possible by lowering the health, and reduce spending on ineffective, wasteful care. administrative costs of insurance in the U.S. If these Achieving benchmark levels of performance, even costs were the same average share of health spending as among a subset of indicators, would yield considerable in three European countries that rely on mixed private/ gains. For example, if the U.S. reduced its mortality rate public insurance, the U.S. could save up to $51 billion from causes amenable to health care to international each year—or more than half the cost to cover the entire benchmarks, approximately 101,000 deaths could be uninsured population. Lowering administrative costs to prevented annually. The National Committee for Quality benchmark levels achieved in the best countries could Assurance (NCQA) estimates that improving national save up to $102 billion. rates of controlling hypertension and diabetes to those The wide variation in costs points to opportunities achieved by the top group of health plans could save for net national gains from the provision of more efficient 16,000 to 39,000 lives each year.83 Some of the potential care. If annual per person costs for Medicare in higher-cost improvements may affect the same individuals. Still, these states came down to median rates or those achieved in the estimates serve as compelling evidence of the human and lowest-quartile states, the nation would save $22 billion economic costs of poor performance. to $38 billion per year.85 Estimated savings from these In addition to reducing mortality, health performance selected improvements toward more effective, timely, improvement has the potential to improve quality of life and coordinated care are only a fraction of more than $2 from preventing disease, disability, and complications. trillion in health spending in 2006. Yet, taken together, Increasing adult preventive care to reach 80 percent of they offer targets to reduce costs and improve value. the population translates to about 70 million more adults Moreover, the nation would gain from improved reaping the benefits of disease prevention and early productivity. The Institute of Medicine estimates national detection. Likewise, 37 million additional adults would have economic gains of up to $130 billion per year from insuring a regular provider for primary care and specialty referrals. the uninsured.86 A recent update of this analysis estimated Closing gaps between average performance and potential savings of up to $204 billion in 2006.87 achieved benchmarks across quality and access indicators The Scorecard highlights the need for a multifaceted also has the potential to reduce costs. If the nation were approach of mutually supporting policies addressing able to meet the benchmark levels of health system access, quality, and efficiency simultaneously. Starting performance on even a select set of indicators, the nation sooner rather than later has the potential to accumulate could save at least $50 billion to $100 billion per year. into substantial gains over time. A recent study prepared Opportunities for savings come from improving outcomes for the Commission on a High Performance Health System 40 illustrates that it would be possible to save $1.5 trillion in Conversely, there was failure to improve in areas where national health expenditures over 10 years and improve we lack metrics or focused efforts to measure, compare, and value in terms of access, quality, and outcomes through improve at the local or facility level. These areas include strategic options including better information, payment mental health care, primary care, hospital readmission changes, and public health improvements, combined with rates, or adverse drug events. Further, the continued insurance for all.88 In sum, raising levels of performance failure to adopt health information technology makes it to benchmarks offers the potential for significant national difficult to generate the information necessary to document gains in health and value. performance and monitor improvement efforts. Looking across dimensions and trends, the 2008 Scorecard reveals several underlying patterns that have B E T T ER P RIMARY C ARE AND C ARE implications for policy. C OORDINA T ION HOLD P O T EN T IAL F OR IM P RO V ED OU T C OMES A T LOWER C OS T S WHA T RE C EI V ES A T T EN T ION Hospital readmission rates have increased and admissions GE T S IM P RO V ED for conditions sensitive to ambulatory care remain high and Notably, the quality indicators showing significant variable across the country, as do the total costs of caring improvement have all been targets of national and for the chronically ill. Studies indicate that it is possible to collaborative efforts to improve, informed by data prevent hospitalizations or rehospitalizations with better with measurable benchmarks and indicators reached primary care, discharge planning, and follow-up care—a by consensus. These initiatives represent important more integrated, “systems” approach. Following on a breakthroughs from the status quo that can and should recommendation from the Medicare Payment Advisory be emulated in other areas. These positive improvements Commission, the federal government recently proposed in performance demonstrate that change can take place that readmission rates be included in an expanded set rapidly over a relatively short time period when there is of quality indicators that hospitals would be required to leadership and measurement. report to receive the full Medicare payment update.89 Multiple indicators highlight the fact that the U.S. •• Hospital quality indicators for heart attack, has a weak primary care foundation. Investing in primary pneumonia, and heart failure (including provision care with enhanced capacity to provide patients with of discharge instructions) were endorsed by a broad round-the-clock access, manage chronic care, and hospital quality alliance. Improvement on these coordinate care will be key strategies to move to more indicators followed after Medicare made payment organized care systems.90 updates contingent on provision of data and public However, current payment incentives for hospitals, reporting. Hospital quality metrics have also been physicians, and nursing homes do not support coordination the focus of Medicare’s Premier Hospital Quality of care or efficient use of expensive, specialized care.91 Incentive Demonstration and private payer initiatives. Information also fails to flow with patients across sites This combination of public–private collaboration and of care due to lack of health information technology and federal leadership changed the landscape on hospital information exchange systems. These inefficiencies require participation in public reporting, establishing a single the attention of policymakers. consensus set of measures that are now well-accepted for benchmarking and improvement. AIMING HIGHER : T HE C ASE F OR A SYS T EMS •• Hospital standardized mortality ratios were the target A P P ROA C H T O C HANGE of many local and national programs and collaborative In summary, the U.S. health system continues to exhibit initiatives that sought to publicize, implement, and suboptimal performance relative to what is achievable and spread evidence-based care and best practices to to the resources invested. The 2008 Scorecard documents achieve better outcomes. Likewise, chronic disease that there are significant human and economic costs of indicators have been central to NCQA’s monitoring failing to address the problems in our health system. of health plan performance through the Healthcare Recent analysis suggests it could be possible to insure Effectiveness Data and Information Set (HEDIS). everyone and achieve savings with improved value over 41 the next decade.92 It is crucial to recognize that health care access, quality, and efficiency are interrelated. With health care expenditures projected to double to $4 trillion, or 20 percent of national income, and millions more Americans on a path to becoming uninsured or underinsured absent new policies, it is critical to start now on the road to higher performance. Aiming higher will require strategies that address the multiple sources of poor performance. These strategies include: •• universal and well-designed coverage that ensures affordable access and continuity of care; •• incentives aligned to promote higher quality and more efficient care; •• care that is designed and organized around the patient, not providers or insurers; •• widespread implementation of health information technology with information exchange; •• explicit goals to meet and exceed benchmarks and monitor performance; and •• national policies that promote private–public collaboration and high performance.93 As rising costs put family, business, and public budgets under stress, access to care and financial protection are eroding for middle- as well as low-income families. New national policies that take a coherent, whole-system, population view are essential for the nation’s future health and economic security. 42 Notes 1 T. Bodenheimer, “Coordinating Care—A 9 Employee Benefit Research Institute, 2006 20 E. G. Stone, S. C. Morton, M. E. Hulscher et Perilous Journey Through the Health Care Health Confidence Survey: Dissatisfaction al., “Interventions that Increase Use of Adult System,” New England Journal of Medicine, with Health Care System Doubles Since 1998 Immunization and Cancer Screening Services: Mar. 6, 2008 358(10):1064–71. (Washington, D.C: EBRI, Nov. 2006), available A Meta-Analysis,” Annals of Internal Medicine, at http://www.ebri.org/pdf/notespdf/EBRI_ May 7, 2002 136(9):641–51. 2 Medicare Payment Advisory Commission, Notes_11-20061.pdf. “Payment Policy for Inpatient Readmissions,” 21 P. S. Wang, O. Demler, and R. C. Kessler, in Report to the Congress: Promoting Greater 10 Institute of Medicine, Coverage Matters: “Adequacy of Treatment for Serious Mental Efficiency in Medicare (Washington, D.C.: Insurance and Health Care (Washington, Illness in the United States,” American Journal MedPAC, June 2007); D. C. Grabowski, D.C.: National Academies Press, Oct. 2001). of Public Health, Jan. 2002 92(1):92–98. “Medicare and Medicaid: Conflicting 11 K. Davis, C. Schoen, S. C. Schoenbaum, M. 22 National Committee for Quality Assurance, Incentives for Long-Term Care,” Milbank M. Doty, A. L. Holmgren, J. L. Kriss, and The State of Health Care Quality 2007 Quarterly, Dec. 2007 85(4):579–610; R. J. Baron K. K. Shea, Mirror, Mirror on the Wall: An (Washington, D.C.: NCQA, 2007). and C. K. Cassel, “21st Century Primary Care: International Update on the Comparative New Physician Roles Need New Payment 23 Ibid. Performance of American Health Care (New Models,” Journal of the American Medical York: The Commonwealth Fund, May 2007); 24 A. Shih and S. C. Schoenbaum, Measuring Association, Apr. 2, 2008 299(13):1595–97. C. Angrisano, D. Farrell, B. Kocher et al., Hospital Performance: The Importance 3 C. Schoen, S. Guterman, A. Shih, J. Lau, S. Accounting for the Cost of Health Care in of Process Measures (New York: The Kasimow, A. K. Gauthier, and K. Davis, the United States (San Francisco: McKinsey Commonwealth Fund Commission on a High Bending the Curve: Options for Achieving Global Institute, Jan. 2007). Performance Health System, July 2007). See Savings and Improving Value in U.S. Health the Scorecard Chartpack for results using an 12 The Commonwealth Fund Commission on a Spending (New York: The Commonwealth expanded set of measures. High Performance Health System, Why Not Fund Commission on a High Performance the Best? Results from a National Scorecard 25 A. K. Jha, E. J. Orav, Z. Li et al., “The Inverse Health System, Dec. 2007). on U.S. Health System Performance (New Relationship Between Mortality Rates and 4 The Commonwealth Fund Commission on a York: The Commonwealth Fund, Sept. Performance in the Hospital Quality Alliance High Performance Health System, A High 2006). Measures,” Health Affairs, July/Aug. 2007 Performance Health System for the United 26(4):1104–10. 13We also assessed annual rates of change to States: An Ambitious Agenda for the Next account for the different time periods in the 26 A measure of patient-centered medical homes President (New York: The Commonwealth updates, with a threshold of 2.5 percent per for children could not yet be updated. Fund, Nov. 2007). year. This method yielded similar results to 27 B. Starfield, Primary Care: Balancing Health 5 J. Holahan and A. Cook, “The U.S. Economy the 5 percent threshold of change. Needs, Services, and Technology (New York: and Changes in Health Insurance Coverage, 14 E. Nolte and C. M. McKee, “Measuring the Oxford University Press, 1998). 2000–2006,” Health Affairs Web Exclusive Health of Nations: Updating an Earlier (Feb. 20, 2008):w135–w144; C. DeNavas-Walt, 28 E. A. Coleman, “Falling Through the Cracks: Analysis,” Health Affairs, Jan./Feb. 2008 B. D. Proctor, and C. H. Lee, Income, Poverty, Challenges and Opportunities for Improving 27(1):58–71. and Health Insurance Coverage in the United Transitional Care for Persons with Continuous States: 2005 (Washington, D.C.: U.S. Census 15 E. Eckholm, “In Turnabout, Infant Deaths Climb Complex Care Needs,” Journal of the Bureau, Aug. 2006). in South,” New York Times, Apr. 22, 2007. American Geriatrics Society, Apr. 2003 51(4):549–55. 6 Kaiser Family Foundation, “Wages and Benefits: 16 C. L. Peterson and R. Burton, U.S. Health Care A Long-Term View” (Washington, D.C.: KFF, Spending: Comparison with Other OECD 29W. D. Klinkenberg and R. J. Calsyn, “Predictors Feb. 2008), available at http://www.kff.org/ Countries, Pub. no. RL34175 (Washington, of Psychiatric Hospitalization: A Multivariate insurance/snapshot/chcm012808oth.cfm. D.C.: Congressional Research Service, Sept. Analysis,” Administration and Policy in 2007). Mental Health, Mar. 1998 25(4):403–10; W. 7 S. Keehan, A. Sisko, C. Truffer et al., “Health D. Klinkenberg and R. J. Calsyn, “Predictors Spending Projections Through 2017: The 17 M. P. Heron, D. L. Hoyert, J. Xu et al., “Deaths: of Receipt of Aftercare and Recidivism Baby-Boom Generation Is Coming to Preliminary Data for 2006,” National Vital Among Persons with Severe Mental Illness: Medicare,” Health Affairs Web Exclusive (Feb. Statistics Reports, vol. 56, no. 16 (Hyattsville, A Review,” Psychiatric Services, May 1996 26, 2008):w145–w155; P. R. Orszag and P. Ellis, Md.: National Center for Health Statistics, 47(5):487–96; S. C. Schoenbaum, D. Cookson, “Addressing Rising Health Care Costs—A June 11, 2008); World Health Statistics and S. Stelovich, “Postdischarge Follow-Up View from the Congressional Budget Office,” 2008 (Geneva, Switzerland: World Health of Psychiatric Inpatients and Readmission in New England Journal of Medicine, Nov. 8, Organization, 2008). an HMO Setting,” Psychiatric Services, Sept. 2007 357(19):1885–87. 18 K. E. Thorpe, D. H. Howard, and K. 1995 46(9):943–45. 8 R. Mangione-Smith, A. H. DeCristofaro, C. M. Galactionova, “Differences in Disease 30 R. T. Konetzka, W. Spector, and M. R. Setodji et al., “The Quality of Ambulatory Prevalence as a Source of the U.S.–European Limcangco, “Reducing Hospitalizations Care Delivered to Children in the United Health Care Spending Gap,” Health Affairs from Long-Term Care Settings,” Medical States,” New England Journal of Medicine, Web Exclusive (Oct. 2, 2007):w678–w686; Care Research and Review, Feb. 2008 Oct. 11, 2007 357(15):1515–23; E. A. McGlynn, J. M. McWilliams, E. Meara, A. M. Zaslavsky 65(1):40–66. S. M. Asch, J. Adams et al., “The Quality of et al., “Use of Health Services by Previously Health Care Delivered to Adults in the United Uninsured Medicare Beneficiaries,” New 31 Medicare Payment Advisory Commission, States,” New England Journal of Medicine, England Journal of Medicine, July 12, 2007 “Skilled Nursing Facility Services,” in Report June 26, 2003 348(26):2635–45; Institute of 357(2):143–53. to the Congress: Medicare Payment Policy Medicine, Crossing the Quality Chasm: A (Washington, D.C.: MedPAC, 2007). 19 A straight average across all indicators in New Health System for the 21st Century quality rather than across sub-dimensions 32 Grabowski, “Medicare and Medicaid: (Washington, D.C.: National Academies Press, yields the same score. Conflicting Incentives,” 2007. Jan. 2001). 43 33 J. Lynn, J. West, S. Hausmann et al., 45 Underinsured is defined as having health care 57 E. S. Fisher, D. E. Wennberg, T. A. Stukel et “Collaborative Clinical Quality Improvement expenses exceeding 10 percent of family al., “The Implications of Regional Variations for Pressure Ulcers in Nursing Homes,” Journal income (or 5 percent for those with incomes in Medicare Spending. Part 1: The Content, of the American Geriatrics Society, Oct. 2007 below 200 percent of the federal poverty Quality, and Accessibility of Care,” Annals of 55(10):1663–69. level) or deductibles that alone constituted Internal Medicine, Feb. 18, 2003 138(4):273–87; 5 percent of income. E. S. Fisher, D. E. Wennberg, T. A. Stukel et al., 34 B. Jarman, A. Bottle, P. Aylin et al., “Monitoring “The Implications of Regional Variations in Changes in Hospital Standardised Mortality 46 Kaiser Family Foundation and Health Research Medicare Spending. Part 2: Health Outcomes Ratios,” British Medical Journal, Feb. 12, 2005 and Educational Trust, Employer Health and Satisfaction with Care,” Annals of Internal 330(7487):329. Benefits 2007 Annual Survey (Washington, Medicine, Feb. 18, 2003 138(4):288–98. D.C./Chicago: KFF/HRET, Sept. 2007), 35 Institute for Healthcare Improvement, available at http://www.kff.org/insurance/7672/ 58 Calculated based on data in A. Catlin, C. Move Your Dot: Measuring, Evaluating, upload/76723.pdf. Cowan, M. Hartman et al., “National Health and Reducing Hospital Mortality Rates Spending in 2006: A Year of Change for (Cambridge, Mass.: IHI, 2003). 47 S. R. Collins, C. Schoen, K. Davis, A. K. Gauthier, Prescription Drugs,” Health Affairs, Jan./Feb. and S. C. Schoenbaum, A Roadmap to Health 36 D. McCarthy, Case Study: The National Surgical 2008 27(1):14–29. Insurance for All: Principles for Reform (New Quality Improvement Program (New York: York: The Commonwealth Fund Commission 59 A. C. Beal, M. M. Doty, S. E. Hernandez, K. K. The Commonwealth Fund, May 2005). on a High Performance Health System, Oct. Shea, and K. Davis, Closing the Divide: How 37 J. Garbutt, A. D. Waterman, J. M. Kapp et 2007). Medical Homes Promote Equity in Health al., “Lost Opportunities: How Physicians Care—Results from the Commonwealth Fund 48 S. Dorn, Uninsured and Dying Because of It: Communicate About Medical Errors,” Health 2006 Health Care Quality Survey (New York: Updating the Institute of Medicine Analysis Affairs, Jan./Feb. 2008 27(1):246–55. The Commonwealth Fund, June 2007). on the Impact of Uninsurance on Mortality 38 P. Pronovost, D. Needham, S. Berenholtz et (Washington, D.C.: Urban Institute, Jan. 60J. D. Albano, E. Ward, A. Jemal et al., al., “An Intervention to Decrease Catheter- 2008). “Cancer Mortality in the United States by Related Bloodstream Infections in the ICU,” Education Level and Race,” Journal of the 49 M. V. Pauly and J. A. Pagan, “Spillovers and New England Journal of Medicine, Dec. National Cancer Institute, Sept. 19, 2007 Vulnerability: The Case of Community 28, 2006 355(26):2725–32. See Scorecard 99(18):1384–94. Uninsurance,” Health Affairs, Sept./Oct. Chartpack. 2007 26(5):1304–14; J. C. Cantor, C. Schoen, D. 61 A. Jemal, E. Ward, R. N. Anderson et al., 39 M. Stewart, L. Meredith, J. B. Brown et al., Belloff, S. K. H. How, and D. McCarthy, Aiming “Widening of Socioeconomic Inequalities “The Influence of Older Patient–Physician Higher: Results from a State Scorecard on in U.S. Death Rates, 1993–2001,” PLoS ONE, Communication on Health and Health-Related Health System Performance (New York: The May 2008 3(5):e2181. Outcomes,” Clinics in Geriatric Medicine, Feb. Commonwealth Fund Commission on a High 62 Dorn, Uninsured and Dying Because of It, 2000 16(1):25–36; M. A. Stewart, “Effective Performance Health System, June 2007). 2008. Physician–Patient Communication and 50 Collins, Schoen, Davis et al., Roadmap to Health Outcomes: A Review,” Canadian 63 E. Ward, M. Halpern, N. Schrag et al., Health Insurance, 2007. Medical Association Journal, May 1, 1995 “Association of Insurance with Cancer 152(9):1423–33. 51 E. S. Fisher and H. G. Welch, “Avoiding the Care Utilization and Outcomes,” CA: A Unintended Consequences of Growth in Cancer Journal for Clinicians, Jan./Feb. 2008 40 R. A. Lowe, A. R. Localio, D. F. Schwarz et Medical Care: How Might More Be Worse?” 58(1):9–31; M. T. Halpern, E. M. Ward, A. L. al., “Association Between Primary Care Journal of the American Medical Association, Pavluck et al., “Association of Insurance Status Practice Characteristics and Emergency Feb. 3, 1999 281(5):446–53. and Ethnicity with Cancer Stage at Diagnosis Department Use in a Medicaid Managed for 12 Cancer Sites: A Retrospective Analysis,” Care Organization,” Medical Care, Aug. 2005 52 A. Tynan, R. A. Berenson, and J. B. Christianson, Lancet Oncology, Mar. 2008 9(3):222–31. 43(8):792–800. “Health Plans Target Advanced Imaging Services” Issue brief no. 118 (Washington, 64 B. Starfield, L . Shi, and J. Macinko, 41 G. Makoul and R. H. Curry, “The Value of D.C.: Center for Studying Health System “Contribution of Primary Care to Health Assessing and Addressing Communication Change, Feb. 2008), available at http://www. Systems and Health,” Milbank Quarterly, Skills,” Journal of the American Medical hschange.com/CONTENT/968/968.pdf. 2005 83(3):457–502; K. Baicker and A. Association, Sept. 5, 2007 298(9):1057–59; Chandra, “Medicare Spending, the Physician L. L. Berry, K. Seiders, and S. S. Wilder, 53 Analysis for the Scorecard by G. F. Anderson, Workforce, and Beneficiaries’ Quality of “Innovations in Access to Care: A Patient- Johns Hopkins Bloomberg School of Public Care,” Health Affairs Web Exclusive (Apr. 7, Centered Approach,” Annals of Internal Health. See Appendix B for list of conditions 2004):w184–w197. Medicine, Oct. 7, 2003 139(7):568–74; D. used to assess readmission rates. M. Post, D. J. Cegala, and W. F. Miser, “The 65 J. Macinko, B. Starfield, and L. Shi, “The 54 MedPAC, “Payment Policy for Inpatient Other Half of the Whole: Teaching Patients Contribution of Primary Care Systems to Readmissions,” 2007. to Communicate with Physicians,” Family Health Outcomes Within Organization for Medicine, May 2002 34(5):344–52. 55 Ibid. Economic Cooperation and Development (OECD) Countries, 1970–1998,” Health 42 J. S. Banthin, P. Cunningham, and D. M. Bernard, 56 G. F. Anderson and J. Horvath, “Chronic Services Research, June 2003 38(3):831–65. “Financial Burden of Health Care, 2001–2004,” Conditions: Making the Case for Ongoing Health Affairs, Jan./Feb. 2008 27(1):188–95. Care” (Baltimore, Md.: Partnership for 66 T. Bodenheimer, “Primary Care—Will It Solutions, Dec. 2002), available at http:// Survive?” New England Journal of Medicine, 43 The scoring remains basically the same if www.partnershipforsolutions.org/DMS/files/ Aug. 31, 2006 355(9):861–64. averaged across indicators instead of the chronicbook2002.pdf. two sub-dimensions. 67 H. T. Tu and A. S. O’Malley, Exodus of Male Physicians from Primary Care Drives Shift 44 Analysis of the U.S. Census Bureau’s to Specialty Practice. Tracking Report No. Current Population Survey. See Scorecard 17 (Washington, D.C.: Center for Studying Chartpack. Health System Change, June 2007); T. Bodenheimer, R. A. Berenson, and P. Rudolf, “The Primary Care–Specialty Income Gap: Why It Matters,” Annals of Internal Medicine, Feb. 20, 2007 146(4):301–06. 44 68 American Academy of Family Physicians, 79L. A. Grant, Culture Change in a For-Profit National Residency Matching Program Nursing Home Chain: An Evaluation (New Summary and Analysis: Comparison of York: The Commonwealth Fund, Feb. 2008); Primary Care Positions Filled with U.S. Seniors R. I. Stone, S. C. Reinhard, B. Bowers et al., in March (1997–2008), available at http:// Evaluation of the Wellspring Model for www.aafp.org/online/en/home/residents/ Improving Nursing Home Quality (New York: match/graph5.html. The Commonwealth Fund, Aug. 2002). 69 Bodenheimer, “Primary Care—Will It 80 See Scorecard Chartpack for data from the Survive?” 2006. Commonwealth Fund National Survey of Nursing Homes. 70L. H. Aiken, S. P. Clarke, and D. M. Sloane, “Hospital Staffing, Organization, and 81 Alliance for Health Reform, Change the Quality of Care: Cross-National Findings,” Nursing Home Culture (Washington, D.C.: International Journal for Quality in Health Alliance for Health Reform, Mar. 2008). Care, Feb. 2002 14(1):5–13; D. C. Vahey, L. H. 82 Schoen, Guterman, Shih et al., Bending the Aiken, D. M. Sloane et al., “Nurse Burnout Curve, 2007. and Patient Satisfaction,” Medical Care, Feb. 2004 42(2 Suppl):II57–II66; D. E. Yeatts and 83 NCQA, The State of Health Care Quality, C. M. Cready, “Consequences of Empowered 2007. CNA Teams in Nursing Home Settings: A 84 Ibid. Longitudinal Assessment,” Gerontologist, June 2007 47(3):323–39. 85 Cantor, Schoen, Belloff et al., Aiming Higher, 2007. 71 Government Accountability Office, Nursing Workforce: Recruitment and Retention 86 Institute of Medicine, Hidden Costs, Value of Nurses and Nurse Aides Is a Growing Lost: Uninsurance in America (Washington, Concern (Washington, D.C.: GAO, 2001); D.C.: National Academies Press, June 2003). L. J. Hayes, L. O’Brien-Pallas, C. Duffield et 87 S. Axeen and E. Carpenter, Cost of Failure: al., “Nurse Turnover: A Literature Review,” The Economic Losses of the Uninsured International Journal of Nursing Studies, Feb. (Washington, D.C.: New America Foundation, 2006 43(2):237–63. Mar. 2008). 72 See Scorecard Chartpack for national 88 Schoen, Guterman, Shih et al., Bending the distribution of scores on Press Ganey’s Curve, 2007. (Hospital) Employee Perspectives Survey and MyInnerView’s National Survey of Nursing 89 Centers for Medicare and Medicaid Services, Home Workforce Satisfaction. “CMS Proposes to Expand Quality Program for Hospital Inpatient Services in 2009” 73 R. L. Kane, T. A. Shamliyan, C. Mueller et al., (Washington, D.C: CMS Office of Public “The Association of Registered Nurse Staffing Affairs, Apr. 14, 2008). Levels and Patient Outcomes: Systematic Review and Meta-Analysis,” Medical Care, 9 0 Bodenheimer, “Coordinating Care—Perilous Dec. 2007 45(12):1195–204; Abt Associates, Journey,” 2008. Appropriateness of Minimum Nurse Staffing 91 MedPAC, “Payment Policy for Inpatient Ratios in Nursing Homes: Phase II Final Report Readmissions” 2007; Grabowski, “Medicare (Baltimore, Md.: Centers for Medicare and and Medicaid: Conflicting Incentives,” 2007; Medicaid Services, 2001). Baron and Cassel, “21st Century Primary Care,” 74 Chartbook authors’ analysis of data on 2008. “Average Full-time Equivalent Direct Care 92 Schoen, Guterman, Shih et al., Bending the Staff Per Facility by State, CMS OSCAR Data Curve, 2007. Current Surveys, June 2007” compiled by the American Health Care Association. 93 Commonwealth Fund Commission, Ambitious Agenda, 2007. 75 Institute of Medicine, To Err Is Human: Building a Safer Health System (Washington, D.C.: National Academies Press, Nov. 1999). 7 6 K. E. Weick, “Organizational Culture as a Source of High Reliability,” California Management Review, 1987 29(2):112–27. 7 7 See Scorecard Chartpack for data from the Agency for Healthcare Research and Quality Hospital Survey on Patient Safety Culture. A subset of attributes in the survey are discussed and displayed on the chart. 78 D. McCarthy and D. Blumenthal, “Stories from the Sharp End: Case Studies in Safety Improvement,” Milbank Quarterly, Mar. 2006 84(1):165–200. 45 Appendices Appendix A. Indicator, Scoring, and Dimensions Tables Tab l e 1 Changes in Indicator Scores and Rates, 2008 Scorecard Compared with 2006 Scorecard Tab l e 2 Performance Indicators—Healthy Lives Tab l e 3 Performance Indicators—Quality Tab l e 4 Performance Indicators—Quality (continued) Tab l e 5 Performance Indicators—Access Tab l e 6 Performance Indicators—Efficiency Tab l e 7 Performance Indicators—Equity Appendix B. Technical Notes: Scorecard Data Years, Databases, and Sources 47 appendix a. Table 1: Changes in indicator scores and rates, 2008 scorecard Compared with 2006 scorecard Total healthy lives Quality access effi ciency Count Percent Count Percent Count Percent Count Percent Count Percent r aT i o s C o r e s 37 100% 5 100% 19 100% 5 100% 8 100% Score Improved 13 35% 1 20% 7 37% 0 0% 5 63% Score Declined 15 41% 2 40% 7 37% 4 80% 2 25% No Change 5 14% 0 0% 3 16% 1 20% 1 13% Not Updated 4 11% 2 40% 2 11% 0 0% 0 0% u P daT e d i n d i C aTo r s * 53 100% 3 100% 27 100% 5 100% 18 100% National Average Improved (>5%) 16 30% 0 0% 12 44% 0 0% 4 22% National Average Declined (>5%) 15 28% 1 33% 6 22% 3 60% 5 28% little/No Change 22 42% 2 67% 9 33% 2 40% 9 50% * Counts include all indicator and underlying data components with updated national data only. Eight indicators/subcomponents could not be updated. Source: Commonwealth Fund National Scorecard on U.S. Health System Performance, 2008 appendix a. Table 2: Performance indicators—healthy lives score: ratio of 2006 scorecard 2008 scorecard u.s. to Benchmark National Range of Performance National Range of Performance 2006 2008 dimension and indicator Rate (Bottom Group–Top Group) Rate (Bottom Group–Top Group) Scorecard Scorecard Mortality amenable to health care, 115 130–80 110 106–69 70 63 Deaths per 100,000 populationa Infant mortality, Deaths per 1,000 live birthsb 7.0 9.9–4.8 6.8 10.1–4.7 68 69 Healthy life expectancy at age 60, 87 87* years (Average of 2 ratios): Men c 15.3 14.4–17.4 * 88 88* Women c 17.9 17.2–20.8 * 86 86* Adults under 65 limited in any activities because of 14.9 20.1–11.5 17.5 23.4–13.2 77 66 physical, mental, or emotional problems, percentb Children missed 11 or more school days 5.2 8.1–3.8 * 73 73* due to illness or injury, percentb h e a lT h y l i V e s d i m e n s i o n s C o r e 75 72 Sources: See Appendix B. Notes: Ranges of performance show the rates for the bottom (worst) and top (best) group as footnoted by indicator. In 2006, benchmark is the top group rate at that period; in 2008, benchmark is the best top group rate from either period. Underlined indicator scores were used to determine the dimension score. * Indicates no updated data available; used baseline score from 2006. a Average bottom or top three of 19 countries. b Average bottom or top 10 percent of states. c Average bottom or top three of 23 countries. 48 appendix a. Table 3: Performance indicators—Quality score: ratio of 2006 scorecard 2008 scorecard u.s. to Benchmark National Range of Performance National Range of Performance 2006 2008 dimension and indicator Rate (%) (Bottom Group–Top Group) Rate (%) (Bottom Group–Top Group) Scorecard Scorecard eFFeCTiVe Care sCore 74 78 Adults received recommended screening and preventive carea 49 31–52 50 32–53 61 62 Children received recommended immunizations and 85 86 preventive care (Average of 2 ratios): Received all recommended doses of fi ve key vaccinesb 79 71–89 81 72–86 89 91 Received both medical and dental preventive care visits b 59 48–73 * 81 81* Needed mental health care and received treatment 76 76 (Average of 2 ratios): Adults with major depressive episode who 65 41–83 69 50–87 79 80 received treatmentc Children b 59 47–74 * 73 73* Chronic disease under control (Average of 2 ratios): 65 76 Adults with diagnosed diabetes whose HbA1c level <9% d 79 23–89 88 30–88 89 98 Adults with hypertension whose blood pressure 31 48–75 41 39–68 41 54 <140/90 mmHgd Hospitalized patients received recommended care for 84 75–91 90 78–95 84 90 heart att ack, heart failure, and pneumonia (composite) e C o o r d i n aT e d C a r e s C o r e 72 71 Adults under 65 with accessible primary care provider f 66 38–84 65 37–85 79 76 Children with a medical homeb 46 38–60 * 77 77* Care coordination at hospital discharge (Average of 3 ratios): 70 74 Hospitalized patients with new Rx: 67 67–86 * 78 78* Medications were reviewed at discharge g Heart failure patients received written instructions 50 9–87 68 36–94 58 72 at dischargeh Follow-up within 30 days after hospitalization for 74 72 mental health disorder i (Average of health plans): Private plansd 76 65–86 76 63–88 88 87 Medicare plans d 61 39–80 56 29–81 70 64 Medicaid plansd 54 22–81 58 17–80 63 66 Nursing homes: hospital admissions and readmissions 71 65 among residents (Average of 2 ratios): Hospital admissions j 17 26–11 19 27–12 63 56 Readmissions j 17 21–13 18 22–15 80 73 Home health: hospital admissions k 28 47–17 28 48–19 62 62 Sources: See Appendix B. Notes: Ranges of performance show the rates for the bottom (worst) and top (best) group as footnoted by indicator. In 2006, benchmark is the top group rate from that period; in 2008, benchmark is the best top group rate from either period. Underlined indicator scores were used to determine the dimension score. * Indicates no updated data available; used baseline score from 2006. a Uninsured or insured all year. Benchmark is target rate at 80. b Average bottom or top 10 percent of states. c Uninsured or insured. d 10th or 90th percentile health plans. e 10th or 90th percentile hospitals. Benchmark is top hospital rate at 100. f Uninsured adults under age 65 or high-income elderly. g Worst or best of six countries. h 10th or 90th percentile hospitals. i Average of National Committee for Quality Assurance health plans; no national data available. Benchmark is 90th percentile private plans. j 90th or 10th percentile states. k Average bottom or top 25 percent of agencies 49 appendix a. Table 4: Performance indicators—Quality (continued) score: ratio of 2006 scorecard 2008 scorecard u.s. to Benchmark National Range of Performance National Range of Performance 2006 2008 dimension and indicator Rate (%) (Bottom Group–Top Group) Rate (%) (Bottom Group–Top Group) Scorecard Scorecard sa F e C a r e s C o r e 69 68 Patients reported medical, medication, or lab test errora 34 34–22 32 32–19 65 59 Unsafe drug use (Average of 3 ratios): 60 55 Ambulatory care visits for treating adverse drug 15 19–11 20 28–16 71 55 effects, per 1,000 population per yearb Children prescribed antibiotics for throat infection 43 75–12 35 74–14 27 33 without a “strep” testc Elderly used 1 of 33 inappropriate drugsb 18 20–15 17 20–13 83 76 Nursing home residents with pressure sores (Average of 67 66 2 ratios): High-risk residents d 13 18–8 13 17–7 61 59 Short-stay residentsd 19 24–14 17 23–12 73 73 Hospital-standardized mortality ratios, actual to 101 118–85 82 89–74 84 90 expected deathse PaT i e n T- C e n T e r e d, T i m e ly C a r e s C o r e 72 69 Ability to see doctor on same/next day when sick or 47 36–81 46 32–74 58 57 need medical attentiona Very/somewhat easy to get care after hours without 38 38–72 25 25–49 53 35 going to the emergency rooma Doctor–patient communication: always listened, 54 55–74 57 59–75 74 75 explained, showed respect, spent enough timef Adults with chronic conditions given self-management plan a 58 37–65 * 89 89* Patient-centered hospital care (Avg. 3 ratios): 87 87 Staff always managed pain well g 70 61–79 67 60–75 89 90 Staff always responded when needed help to get to 63 52–74 60 48–72 86 83 the bathroom or pressed call button g Staff always explained medicines and side effectsg 60 49–70 58 49–66 86 87 Q ua l i T y d i m e n s i o n s C o r e 72 71 Sources: See Appendix B. Notes: Ranges of performance show the rates for the bottom (worst) and top (best) group as footnoted by indicator. In 2006, benchmark is the top group rate from that period; in 2008, benchmark is the best top group rate from either period. Exception is patient-centered hospital care; in 2008, benchmark is the top group rate from that period. Underlined indicator scores were used to determine the dimension score. * Indicates no updated data available; used baseline score from 2006. a In 2006, worst or best of six countries; in 2008, worst or best of seven countries. b Worst or best of four regions. c 90th or 10th percentile health plans. d Average bottom or top 10 percent of states. e Average bottom or top 10 percent of hospitals. f 10th or 90th percentile health plans. g 10th or 90th percentile hospitals. 50 appendix a. Table 5: Performance indicators—access score: ratio of 2006 scorecard 2008 scorecard u.s. to Benchmark National Range of Performance National Range of Performance 2006 2008 dimension and indicator Rate (%) (Bottom Group–Top Group) Rate (%) (Bottom Group–Top Group) Scorecard Scorecard Pa rT i C i PaT i o n s C o r e 65 62 Adults under 65 insured all year, not underinsured a 65 32–83 58 28–73 65 58 Adults with no access problem due to costsb 60 60–91 63 63–95 66 66 a F F o r da B l e C a r e d i m e n s i o n s C o r e 69 54 Families spending <10% of income or <5% of income, if 81 56–95 77 56–92 81 77 low income, on OOP medical costs and premiumsc Population under 65 living in states where premiums for employer-sponsored health coverage are <15% of 58 NA 25 NA 58 25 under-65 median household income Adults under 65 with no medical bill problems or 66 53–84 59 44–79 66 59 medical debtd aC C e s s d i m e n s i o n s C o r e 67 58 Sources: See Appendix Table B. Notes: Ranges of performance shows the rates for the bottom (worst) and top (best) group as footnoted by indicator. Benchmark is 100 percent of the U.S. population meeting each threshold. Exception is access problems due to cost; in 2006, benchmark is top group rate from that period, and in 2008, benchmark is best top group rate from either period. Underlined indicator scores were used to determine the dimension score. OOP is out-of-pocket. NA Indicates not applicable. a less than 200 percent of the federal poverty level or 200 percent or more of poverty. b In 2006, worst or best of five countries; in 2008, worst or best of seven countries. c less than 100 percent of the federal poverty level or 400 percent or more of poverty. d less than 200 percent of the federal poverty level or 400 percent of more of poverty. 51 appendix a. Table 6: Performance indicators—efficiency score: ratio of u.s. 2006 scorecard 2008 scorecard to Benchmark National Range of Performance National Range of Performance 2006 2008 dimension and indicator Rate (Bottom Group–Top Group) Rate (Bottom Group–Top Group) Scorecard Scorecard Potential overuse or waste (Average of 3 ratios): 48 41 Duplicate medical tests: doctor ordered test that 18 20–6 20 20–4 33 20 had already been done, percenta Tests results or records not available at time of 23 23–11 22 22–9 48 41 appointment, percenta Received imaging study for acute low back pain with 62 62 no risk factors, percentb (Average of health plans): Private plansc 25 33–18 26 35–19 58 56 Medicaid plansc 22 28–15 22 29–15 66 67 Went to emergency room for condition that could have 26 26–6 21 21–6 23 29 been treated by regular doctor, percenta Hospital admissions for ACS conditions (Average of 2 56 56 ratios): National ACS admissions, per 100,000 population 48 45 (Average of 3 conditions): Heart failure d 498 631–258 476 634–246 52 52 Diabetes (composite) d 241 299–137 240 293–126 57 52 Pediatric asthma d 178 242–62 156 230–49 35 31 Medicare ACS admissions, per 10,000 benefi ciariese 771 1043–499 700 926–465 65 66 Medicare hospital 30-day readmission rates, percente 18 22–14 18 21–14 75 76 Medicare annual costs of care and mortality for heart 88 89 attacks, hip fractures, or colon cancer (Average of 2 ratios): Resource costs, annual Part A and Part B $e $26,829 $29,047–$23,314 $28,011 $30,263–$24,906 87 89 1-year mortality rate, percente 30 32–27 30 33–27 90 89 Medicare annual costs of care for chronic diseases: Diabetes, heart failure, COPD, Part A and Part B $ 68 71 (Average of 4 ratios): All three conditionse $31,792 $43,973–$20,960 $38,004 $53,019–$25,732 66 68 Diabetes + heart failure e $18,461 $27,310–$12,747 $23,056 $32,199–$16,144 69 70 Diabetes + COPD e $13,188 $18,024–$8,872 $15,367 $20,062–$11,317 67 74 Heart failure + COPD e $22,415 $32,732–$15,355 $27,498 $37,450–$19,787 69 72 Health insurance administration as percent of national 7.4 6.8–2.4 7.5 6.7–2.3 33 31 health expenditures f Physicians using electronic medical records, percent g 17 7–80 28 23–98 21 29 eFFiCienCy dimension sCore 52 53 Sources: See Appendix B. Notes: Ranges of performance show the rates for the bottom (worst) and top (best) group as footnoted by indicator. In 2006, benchmark is the top group rate from that period; in 2008, benchmark is the best top group rate from either period. Exceptions are cost indicators; in 2008, benchmark is the top group rate from that period. Underlined indicator scores were used to determine the dimension score. ACS is ambulatory care–sensitive. COPD is chronic obstructive pulmonary disease. a In 2006, worst or best of six countries. In 2008, worst or best of seven countries. b Average of National Committee for Quality Assurance health plans; no national data available. Benchmark is 10th percentile Medicaid plans. c 90th or 10th percentile health plans. d Average bottom or top 10 percent of states. e 90th or 10th percentile regions. f Average bottom or top three of 11 countries. g In 2006, average bottom or top three of 19 countries; in 2008, best or worst of seven countries. 52 appendix a. Table 7: Performance indicators—equity high income insured Compared Compared with white Compared white Compared with uninsured low income with Black with hispanic 2006 2008 2006 2008 2006 2008 2006 2008 dimension and indicator Scorecard Scorecard Scorecard Scorecard Scorecard Scorecard Scorecard Scorecard h e a lT h y l i V e s s C o r e – – 54 55 77 77 97 97 Infant mortality – – 63 67 42 42 100 100 Adults under 65 limited in any activities because of – – 46 45 100 100 100 100 physical, mental, or emotional problems Children missed 11 or more school days due to illness – – 51 51* 100 100* 100 100* or injury Cancer 5-year survival – – 82 82* 82 82* 97 97* Coronary heart disease and diabetes-related deaths – – 29 29 64 64 86 88 eFFeCTiVe Care sCore 59 57 69 68 80 76 73 70 Older adults did not receive recommended screening 76 71 80 75 85 84 77 80 and preventive care Children did not receive recommended immunizations 57 57* 58 60 77 83 75 81 and preventive care Needed mental health care and did not receive treatment 43 41 76 87 77 70 69 58 Untreated dental caries 45 43 33 32 49 51 50 50 Chronic disease not under control 66 63 93 85 97 73 92 67 Diabetics did not receive HbA1c, retinal, and foot exams 67 67* 72 67 93 97 78 83 sa F e C a r e s C o r e 97 97 94 94 77 77 94 94 Patient reported medical, medication, or lab test error 100 100* 94 94* 67 67* 100 100* AHRQ patient safety indicators 95 94 95 93 84 86 96 95 Nursing home residents with pressure sores – – – – 79 79 87 86 PaT i e n T- C e n T e r e d, T i m e ly C a r e s C o r e 56 56 59 59 72 62 54 64 6+ days to see doctor when sick or need medical 57 67 54 62 58 44 45 57 attention Doctor–patient communication: sometimes/never 55 46 63 57 86 79 63 71 listened, explained, showed respect, spent enough time Co o r d i n aT e d a n d e F F i C i e n T C a r e sCo r e 55 58 63 60 61 73 58 72 Adults without an accessible primary care provider 47 46 68 66 74 77 63 63 Children without a medical home 62 62* 65 65* 78 78* 68 68* Duplicate medical tests: doctor ordered test that had 43 58 53 65 50 100 43 87 already been done Tests results or records not available at time of 58 61 74 61 62 75 46 75 appointment went to ER for condition that could have been treated 67 65 58 50 41 68 65 100 by regular doctor Hospital admissions for ACS conditions – – 50 42 32 33 51 54 aC C e s s s C o r e 57 61 30 32 86 87 82 87 Adults under 65 with time uninsured during the year – – 28 29 75 73 47 48 Adults under 65 with access problems because of costs 47 49 46 43 100 100 88 100 Families spending >10% of income or >5% of income, if 82 94 11 19 94 93 92 100 low income, on OOP medical costs and premiums Adults under 65 with medical bill problems or medical debt 50 54 34 38 75 80 100 100 s C o r e By e Q u i T y g ro u P 65 66 61 61 76 75 77 81 Sources: See Appendix B. ER is emergency room. ACS is ambulatory–care sensitive. OOP is out-of-pocket. * Indicates no updated data available; used baseline score from 2006. – Indicates not scored. 53 Appendix B. Technical Notes: Scorecard Data Years, Databases, and Sources The following list provides additional information for all indicators, including: or researchers who conducted new data analysis. Further descriptions are provided 1) the date for national and benchmark data used in the 2006 and 2008 scorecards; below for select indicators marked by an asterisk. 2) database; and 3) citation for data drawn from published sources, online databases, Year for Year for 2006 2008 Scorecard Scorecard Database Source Notes LONG , HEALT HY & P RODU C T I V E LI V ES 1. Mortality amenable to health care* 1997-1998 2002-2003 WHO mortality files E. Nolte and C.M. McKee, “Measuring the Health of Nations: Updating an Earlier Analysis,” Health Affairs 27, no. 1 (2008): 58-71. 2. Infant mortality 2002 2004 NVSS-I AHRQ , National Healthcare Quality Report, Data Tables Appendix (2005, 2007). 3. Healthy life expectancy at age 60 3.1. Men 2002 Not Updated WHO WHO, The World Health Report 2003: Shaping the Future (Geneva: WHO, 2003). 3.2. Women 2002 Not Updated Same as above. Same as above. 4. Adults under 65 limited in any activities because 2004 2006 BRFSS Analysis by D. Belloff, Rutgers Center for State of physical, mental, or emotional problems Health Policy. 5. Children missed 11 or more school 2003 Not Updated NSCH Retrieved from the Data Resource Center website at days due to illness or injury http://www.nschdata.org. Q UALI T Y 6. Adults received recommended 2002 2005 MEPS Analysis by B. Mahato, Columbia University. screening and preventive care* 7. Children received recommended immunizations and preventive care 7.1. Received all recommended doses of five key vaccines 2003 2006 NIS NCHS, National Immunization Program. 7.2. Received both medical and dental 2003 Not Updated NSCH Retrieved from the Data Resource Center website at preventive care visits http://www.nschdata.org. 8. Needed mental health care and received treatment 8.1. Adults with major depressive episode 2004 2006 NSDUH SAMHSA, Results from the National Survey on Drug who received treatment Use and Health: National Findings (2006, 2007). 8.2. Children 2003 Not Updated NSCH Retrieved from the Data Resource Center at http:// www.nschdata.org. 9. Chronic disease under control 9.1. Adults with diagnosed diabetes whose 1999-2000 2003-2004 NHANES Analysis by J. M. McWilliams, Harvard Medical HbA1c level <9%: national data School. Adults with diagnosed diabetes whose 2004 2006 HEDIS NCQA, HEDIS Audit Means, Percentiles and Ratios HbA1c level <9%: benchmark data (2005, 2007). 9.2. Adults with hypertension whose blood 1999-2000 2003-2004 NHANES Analysis by J. M. McWilliams, Harvard Medical pressure <140/90 mmHg: national data School. Adults with hypertension whose blood 2004 2006 HEDIS NCQA , HEDIS Audit Means, Percentiles and Ratios pressure <140/90 mmHg: benchmark data (2005, 2007). 10. Hospitalized patients received recommended care for 2004 2006 CMS Hospital Compare Analysis by A. Jha and A. Epstein, Harvard School of heart attack, heart failure, and pneumonia (composite)* Public Health. 11. Adults under 65 with accessible primary care provider* 2002 2005 MEPS Analysis by B. Mahato, Columbia University 12. Children with a medical home 2003 Not Updated NSCH Retrieved from the Data Resource Center website at http://www.nschdata.org. 13. Care coordination at hospital discharge 13.1. Hospitalized patients with new Rx: 2005 Not Updated Commonwealth Analysis by authors using survey sample of adults Medications were reviewed at discharge Fund IHP Survey with health problems. 13.2. Heart failure patients received written 2004 2006 CMS Hospital Compare Analysis by A. Jha and A. Epstein, Harvard School of instructions at discharge Public Health. 13.3. Follow-up within 30 days after 2004 2006 HEDIS NCQA , HEDIS Audit Means, Percentiles and Ratios hospitalization for mental health disorder: (2005, 2007). private plans, Medicare, Medicaid 14. Nursing homes: hospital admissions and readmissions among residents 14.1. Hospital admissions 2000 2004 MEDPAR, MDS Analysis by V. Mor, Brown University. 14.2. Readmissions 2000 2004 Same as above. Same as above. 15. Home health: hospital admissions 2003-2004 2006-2007 OASIS 2003–2004 data from K. Pace et al., Acute hospitalization of home health patients report of analyses, literature review, and technical expert panel (2005); 2006–2007 data retrieved from CMS Home Health Compare database at http://www.medicare. gov/HHCompare. 55 Year for Year for 2006 2008 Scorecard Scorecard Database Source Notes 16. Patients reported medical, medication, or lab test error 2005 2007 Commonwealth Analysis by authors using survey sample of adults Fund IHP Survey with health problems. 17. Unsafe drug use 17.1. Ambulatory care visits for treating 2001 2004 NAMCS-NHAMCS Analysis by C. Zhan, AHRQ. adverse drug effects 17.2. Children prescribed antibiotics for throat 1997-2003 2004 NAMCS-NHAMCS Analysis by J. Linder, Brigham and Women’s Hospital. infection without a “strep” test: national data Children prescribed antibiotics for throat 2004 2006 HEDIS NCQA, HEDIS Audit Means, Percentiles and Ratios infection without a “strep” test: benchmark data (2005, 2007). 17.3. Elderly used 1 of 33 inappropriate drugs 2002 2004 MEPS AHRQ , National Healthcare Quality Report, Data Tables Appendix (2005, 2007). 18. Nursing home residents with pressure sores 18.1. High-risk residents 2004 2006 MDS AHRQ , National Healthcare Quality Report, Data Tables Appendix (2005, 2007). 18.2. Short-stay residents 2004 2006 Same as above. Same as above. 19. Hospital-standardized mortality ratios* 2000-2002 2004-2006 Medicare data Analysis by Sir Brian Jarman, Imperial College London, United Kingdom. 20. Ability to see doctor on same/next day 2005 2007 Commonwealth Analysis by authors using survey sample of adults when sick or need medical attention Fund IHP Survey with health problems. 21. Very/somewhat easy to get care after 2005 2007 Commonwealth Analysis by authors using survey sample of adults hours without going to the ER Fund IHP Survey with health problems. 22. Doctor-patient communication: always listened, explained, 2002 2004 MEPS AHRQ , National Healthcare Quality Report, Data showed respect, spent enough time: national data Tables Appendix (2005, 2007). Doctor-patient communication: always listened, explained, 2004 2006 CAHPS Provided by NCQA. showed respect, spent enough time: benchmark data 23. Adults with chronic conditions given self-management plan 2005 Not Updated Commonwealth Analysis by authors using survey sample of adults Fund IHP Survey with health problems. 24. Patient-centered hospital care 24.1. Staff always managed pain well 2005 2007 HCAHPS 2005 data provided by Dale Shaller and AHRQ CAHPS benchmarking database team; 2007 data retrieved from CMS Hospital Compare database at www.hospitalcompare.hhs.gov. 24.2. Staff always responded when needed help to 2005 2007 Same as above. Same as above. get to the bathroom or pressed call button 24.3. Staff always explained medicines and side effects 2005 2007 Same as above. Same as above. AC C ESS 25. Adults under 65 insured all year, not underinsured 2003 2007 Commonwealth Analysis by authors. Fund Biennial Health Insurance Survey 26. Adults with no access problem due to costs 2004 2007 Commonwealth Analysis by authors. Fund IHP Survey 27. Families spending <10% of income or <5% of income, 2001 2005 MEPS Analysis by P. Cunningham, Center for Studying if low income, on OOP medical costs and premiums Health System Change. 28. Population under 65 living in states where premiums 2003 2005 MEPS (premiums), CPS Analysis of CPS by B. Mahato, Columbia University. for employer-sponsored health coverage are (household income) Complete analysis by authors. <15% of under-65 median household income 29. Adults under 65 with no medical bill 2005 2007 Commonwealth Analysis by authors. problems or medical debt Fund Biennial Health Insurance Survey E F F I C IEN C Y 30. Potential overuse or waste 30.1. Duplicate medical tests: doctor ordered 2005 2007 Commonwealth Analysis by authors using survey sample of adults test that had already been done Fund IHP Survey with health problems. 30.2. Tests results or records not available 2005 2007 Commonwealth Analysis by authors using survey sample of adults at time of appointment Fund IHP Survey with health problems. 30.3. Received imaging study for acute low back pain 2004 2006 HEDIS NCQA, HEDIS Audit Means, Percentiles and Ratios with no risk factors: Private plans, Medicaid (2005, 2007). 31. Went to ER for condition that could have 2005 2007 Commonwealth Analysis by authors using survey sample of adults been treated by regular doctor Fund IHP Survey with health problems. 56 Year for Year for 2006 2008 Scorecard Scorecard Database Source Notes 32. Hospital admissions for ACS conditions 32.1. National ACS admissions 32.1a. Heart failure 2002 2004 HCUP AHRQ , National Healthcare Quality Report, Data Tables Appendix (2005, 2007). 32.1b. Diabetes (composite) 2002 2004 Same as above. Same as above. 32.1c. Pediatric asthma 2003 2004 Same as above. Same as above. 32.2. Medicare ACS admissions* 2003 2005 Medicare SAF 5% Analysis by G. Anderson and R.Herbert, Johns Inpatient Data Hopkins Bloomberg School of Public Health. 33. Medicare hospital 30-day readmission rates* 2003 2005 Medicare SAF 5% Analysis by G. Anderson and R.Herbert, Johns Inpatient Data Hopkins Bloomberg School of Public Health. 34. Medicare annual costs of care and mortality for heart attacks, hip fractures, or colon cancer 34.1. Resource costs, annual Part A and Part B 2000-2002 2004 20% national sample of Analysis by E. Fisher, J. Sutherland, and D. Radley, Medicare beneficiaries Dartmouth Medical School 34.2. 1-year mortality rate 2000-2002 2004 Same as above. Same as above. 35. Medicare annual costs of care for chronic diseases: Diabetes, heart failure, COPD 35.1. All three conditions 2001 2005 Medicare SAF 5% Analysis by G. Anderson and R. Herbert, Johns Inpatient Data Hopkins Bloomberg School of Public Health. 35.2. Diabetes + Heart failure 2001 2005 Same as above. Same as above. 35.3. Diabetes + COPD 2001 2005 Same as above. Same as above. 35.4. Heart Failure + COPD 2001 2005 Same as above. Same as above. 36. Health insurance administration as percent 2003 2005 OECD Health of national health expenditures Data 2007 37. Physicians using electronic medical records 2001 2006 Commonwealth Analysis by authors. Fund International Survey of Physicians E Q UI T Y 1. Infant mortality 2002 2004 NVSS-I AHRQ , National Healthcare Disparities Report, Data Appendix Tables (2005, 2007). 2. Adults under 65 limited in any activities because 2004 2006 BRFSS Analysis by D. Belloff, Rutgers Center for State Health of physical, mental, or emotional problems Policy. 3. Children missed 11 or more school 2003 Not Updated NSCH Retrieved from the Data Resource Center website at days due to illness or injury http://www.nschdata.org. 4. Cancer 5-year survival: race/ethnicity data 1988-1997 Not Updated SEER L. Clegg et al., “Cancer Survival among US Whites and Minorities: A SEER Program Population-Based Study,” Arch Intern Med 162, no. 17 (2002): 1985-93 Cancer 5-year survival: income data 1998-1994 Not Updated SEER G. Singh et al., Area Socioeconomic Variations in US Cancer Incidence, Mortality, Stage, Treatment, and Survival 1975–1999 (Bethesda, MD: National Cancer Institute, 2003). 5. Coronary heart disease and diabetes-related deaths 2003 2004 NVSS-M Retrieved from NCHS DATA2010 database at http:// wonder.cdc.gov/data2010. 6. Older adults (age 50 and over) did not receive 2002 2005 MEPS Analysis by B. Mahato, Columbia University. recommended screening and preventive care 7. Children did not receive recommended immunizations and preventive 7.1 Did not receive all recommended 2003 2006 NIS NCHS, National Immunization Program. doses of five key vaccines 7.2 Did not receive both medical and 2003 Not Updated NSCH Retrieved from the Data Resource Center website at dental preventive care visits http://www.nschdata.org. 8. Needed mental healthcare and did not receive treatment 8.1. Adults with major depressive episode 2004 2006 NSDUH SAMHSA, Results from the National Survey on Drug who did not receive treatment Use and Health: National Findings (2006, 2007). 8.2. Children 2003 Not Updated NSCH Retrieved from the Data Resource Center at http:// www.nschdata.org. 9. Untreated dental caries: race/ethnicity data 1999-2002 2001-2004 NHANES NCHS, Health, United States, 2007 (Hyattsville, M.D.: Centers for Disease Control and Prevention, 2007). Untreated dental caries: income and insurance data 1999-2002 2001-2004 NHANES Analysis by J. M. McWilliams, Harvard Medical School. 57 Year for Year for 2006 2008 Scorecard Scorecard Database Source Notes 10. Chronic disease not under control: 10.1 Adults with diagnosed diabetes whose HbA1c 1999-2002 1999-2004 NHANES AHRQ , National Healthcare Quality Report, Data level ≥9%: race/ethnicity and income data Tables Appendix (2005). Analysis updated by J. M. McWilliams, Harvard Medical School. Adults with diagnosed diabetes whose 1988-1994 1999-2004 NHANES Saaddine et al., “A Diabetes Report Card for the HbA1c level ≥9%: insurance data United States: Quality of Care in the 1990s,” Ann Intern Med 136, no 8: 565-74. Analysis updated by J. M. McWilliams, Harvard Medical School. 10.2 Adults with hypertension whose 1999-2002 1999-2004 NHANES AHRQ , National Healthcare Quality Report, Data blood pressure ≥140/90 mmHg Tables Appendix (2005). Analysis updated by J. M. McWilliams, Harvard Medical School. 11. Diabetic adults (age 40 and over) did not 2002 2004 MEPS AHRQ , National Healthcare Quality Report, Data receive HbA1c, retinal, and foot exams Tables Appendix (2005, 2007). 12. Patients reported medical, medication, or lab test error 2005 2007 Commonwealth Analysis by authors using survey sample of general Fund IHP Survey adult population. 13. AHRQ patient safety indicators 13.1 Failure to rescue 2002 2004 HCUP AHRQ , National Healthcare Quality Report, Data Appendix Tables (2005, 2007). AHRQ , National Healthcare Disparities Report, Data Appendix Tables (2005, 2007). 13.2 Decubitus ulcers 2002 2004 Same as above. Same as above. 13.3 Selected infections due to medical care 2002 2004 Same as above. Same as above. 13.4 Postoperative pulmonary embolus 2002 2004 Same as above. Same as above. or deep vein thrombosis 13.5 Postoperative sepsis 2002 2004 Same as above. Same as above. 14. Nursing home residents with pressure sores 14.1. High-risk residents 2004 2005 MDS AHRQ , National Healthcare Disparities Report, Data Tables Appendix (2005, 2007). 14.2. Short-stay residents 2004 2005 Same as above. Same as above. 15. Waited 6 or more days to see doctor when 2005 2007 Commonwealth Analysis by authors using survey sample of general sick or need medical attention Fund IHP Survey adult population. 16. Doctor-patient communication: sometimes/never 2002 2004 MEPS AHRQ , National Healthcare Quality Report, Data listened, explained, showed respect, spent enough time Tables Appendix (2005, 2007). 17. Adults (age 19 and over) without an 2002 2005 MEPS Analysis by B. Mahato, Columbia University accessible primary care provider 18. Children without a medical home 2003 Not Updated NSCH Retrieved from the Data Resource Center website at http://www.nschdata.org. 19. Duplicate medical tests: doctor ordered 2005 2007 Commonwealth Analysis by authors using survey sample general adult test that had already been done Fund IHP Survey population. 20. Tests results or records not available 2005 2007 Commonwealth Analysis by authors using survey sample general adult at time of appointment Fund IHP Survey population. 21. Went to ER for condition that could have 2005 2007 Commonwealth Analysis by authors using survey sample of general been treated by regular doctor Fund IHP Survey adult population. 22. Hospital admissions for ACS conditions 22.1 Congestive heart failure: race/ethnicity data 2002 2005 NHDS AHRQ , National Healthcare Disparities Report, Data Tables Appendix (2005, 2007). Congestive heart failure: income data 2002 2004 HCUP Retrieved from HCUPnet database at http://hcupnet. ahrq.gov. 22.2 Diabetes (composite): race/ethnicity data 2002 2004 HCUP Calculated by authors from AHRQ , National Healthcare Disparities Report, Data Tables Appendix (2005, 2007). Diabetes (composite): income data 2002 2004 HCUP Calculated by authors from HCUPnet database at http://hcupnet.ahrq.gov. 22.3 Pediatric asthma: race/ethnicity data 2003 2004 HCUP AHRQ , National Healthcare Disparities Report, Data Tables Appendix (2007). Pediatric asthma: income data 2003 2004 HCUP AHRQ , National Healthcare Quality Report, Data Tables Appendix (2007). 23. Adults under 65 with time uninsured during the year 2002 2004 MEPS AHRQ , National Healthcare Disparities Report, Data Tables Appendix (2005, 2007). 24. Adults under 65 with access problems because of costs 2005 2007 Commonwealth Analysis by authors. Fund Biennial Health Insurance Survey 25. Families spending >10% of income or >5% of income, if low 2001 2005 MEPS Analysis by P. Cunningham, Center for Studying income, on out-of-pocket medical costs and premiums Health System Change. 26. Adults under 65 with medical bill 2005 2007 Commonwealth Analysis by authors. problems or medical debt Fund Biennial Health Insurance Survey 58 D e f i n iti o n o f Data b a se s Adults received recommended screening and preventive care: Percent of adults 18 or who received seven key screening or preventive services as recommended by the BRFSS =Behavioral Risk Factor Surveillance System U.S. Preventive Services Task Force, including: blood pressure screening within 2 years; CAHPS =Consumer Assessment of Healthcare Providers and System cholesterol screening within 5 years; Pap test within 3 years for women age 18 and CDC =Centers for Disease Control and Prevention older; mammography within 2 years for women age 40 and older; fecal occult blood CPS =Current Population Survey testing (FOBT) within 2 years or colonoscopy/sigmoidoscopy ever for adults age 50 and HCAHPS =Hospital Consumer Assessment of Healthcare Providers and older (either test); and influenza vaccination within past year for adults 65 or older. Systems Survey HCUP =Healthcare Cost and Utilization Project Hospitalized patients received recommended care for heart attack, heart HEDIS =Healthcare Effectiveness Data and Information Set failure, and pneumonia (composite): Proportion of cases where a hospital provided IHP =International Health Policy the recommended process of care for patients with heart attack (acute myocardial MDS =Nursing Home Minimum Data Set infarction), heart failure, and pneumonia for 10 indicators. The composite includes 5 MEDPAR =Medicare Provider Analysis and Review clinical services for heart attack (aspirin within 24 hours before or after arrival at the MEPS =Medical Expenditure Panel Survey hospital and at discharge; beta-blocker within 24 hours after arrival and at discharge; NAMCS-NHAMCS =National Ambulatory Medical Care Survey-National Hospital and angiotensin-converting enzyme (ACE) inhibitor for left ventricular systolic Ambulatory Care Medical Survey dysfunction), 2 for heart failure (assessment of left ventricular function and the use of NCHS =National Center for Health Statistics an ACE inhibitor for left ventricular dysfunction), and 3 for pneumonia (initial antibiotic NHANES =National Health and Nutrition Examination Survey therapy received within four hours of hospital arrival, pneumococcal vaccination, and NHDS =National Hospital Discharge Survey assessment of oxygenation). NIS =National Immunization Survey NSCH =National Survey of Children’s Health Adults under 65 with accessible primary care provider: Percent of adults ages 19 to NSDUH =National Survey on Drug Use and Health 64 that have a usual source of care who provides preventive care (such as general checks OASIS =Outcome and Assessment Information Set ups, examinations, and immunizations), care for new and ongoing health problems, and OECD =Organization for Economic Cooperation and Development referrals to other health professionals when needed and who is easy to get to. NVSS-I =National Vital Statistics System, Linked Birth and Infant Death Data NVSS-M =National Vital Statistics System, Mortality Data Hospital-standardized mortality ratios: Ratio of actual to expected in-hospital SAF =Standard Analytical Files deaths among Medicare beneficiaries diagnosed with conditions accounting for 80 SAMHSA =Substance Abuse and Mental Health Services Administration percent of inpatient mortality. The number of deaths that would be expected is based WHO =World Health Organization on national hospital death rates, stratified by patient age, sex, race, admission source, admission type and length of stay. Expected rates use national hospital deaths in 2000 as the standard. The standardized ratio is further adjusted for community risk factors using regression analysis. D e f i n iti o ns fo r Se le ct I n d i c ato r s Medicare ACS admissions: Hospital admissions of fee-for-service Medicare beneficiaries Mortality amenable to health care: Number of deaths before age 75 per 100,000 age 65 and older for one of 11 ambulatory care-sensitive conditions (AHRQ Prevention population that resulted from causes considered at least partially treatable or preventable Quality Indicators): short-term diabetes complications, long-term diabetes complications, with timely and appropriate medical care (see list). lower extremity amputation among patients with diabetes, asthma, chronic obstructive pulmonary disease, hypertension, congestive heart failure, angina (without a procedure), Cause of deathsAge dehydration, bacterial pneumonia, and urinary tract infection. Intestinal infections 0-14 Tuberculosis 0-74 Medicare hospital 30-day readmission rates: Fee-for-service Medicare beneficiaries Other infections (diphtheria, tetanus, septicaemia, poliomyelitis) 0-74 age 65 and older with initial admissions due to one of 31 select conditions (see list) who Whooping cough 0-14 are readmitted within 30 days following discharge for the initial admission. Measles 1-14 Malignant neoplasm of colon and rectum 0-74 1.Abnormal heartbeat Malignant neoplasm of skin 0-74 2.Chronic obstructive pulmonary disease (COPD) Malignant neoplasm of breast 0-74 3.Congestive heart failure Malignant neoplasm of cervix uteri 0-74 4.Diabetes with amputation Malignant neoplasm of cervix uteri and body of uterus 0-44 5.Diabetes - medical management Malignant neoplasm of testis 0-74 6.Kidney failure Hodgkin’s disease 0-74 7.Kidney and urinary tract infections Leukaemia 0-44 8.Pneumonia - aspiration Diseases of the thyroid 0-74 9.Pneumonia - infectious Diabetes mellitus 0-49 10.Respiratory failure with mechanical ventilation Epilepsy 0-74 11.Respiratory failure without mechanical ventilation Chronic rheumatic heart disease 0-74 12.Stomach and intestinal bleeding Hypertensive disease 0-74 13.Stroke - hemorrhagic Cerebrovascular disease 0-74 14.Stroke - non-hemorrhagic All respiratory diseases (excluding pneumonia and influenza) 1-14 15.Abdominal aortic aneurysm repair Influenza 0-74 16.Gallbladder removal - laparoscopic Pneumonia 0-74 17.Gallbladder removal - open Peptic ulcer 0-74 18.Hip fracture - surgical repair Appendicitis 0-74 19.Hysterectomy - vaginal Abdominal hernia 0-74 20.Removal of blockage of neck vessels Cholelithiasis and cholecystitis 0-74 21.Bronchitis & asthma, complicated DRG096 Nephritis and nephrosis 0-74 22.Bronchitis & asthma, uncomplicated DRG097 Benign prostatic hyperplasia 0-74 23.Hypotension & fainting, complicated DRG141 Maternal death All 24.Chest pain DRG143 Congenital cardiovascular anomalies 0-74 25.Cirrhosis & alcoholic hepatitis DRG202 Perinatal deaths, all causes, excluding stillbirths All 26.Noncancerous pancreatic disorders DRG204 Misadventures to patients during surgical and medical care All 27.Liver disease except cancer, cirrhosis, alcoholic hepatitis, complicated DRG205 Ischaemic heart disease: 50% of mortality rates included 0-74 28.Medical back problems DRG243 29.Surgery for infectious or parasitic disease DRG415 30.Infection after surgery or trauma DRG418 31.Vascular operations except heart, complicated DRG478 59 Further Reading Publications listed below can be found Aiming Higher: Results from a State Scorecard on on The Commonwealth Fund’s Web site at Health System Performance ( June 2007). Joel C. www.commonwealthfund.org. Cantor, Cathy Schoen, Dina Belloff, Sabrina K. H. How, and Douglas McCarthy. The North Dakota Experience: Achieving High-Performance Health Care Through Rural An Analysis of Leading Congressional Health Care Innovation and Cooperation (May 2008). Douglas Bills, 2005–2007: Part I, Insurance Coverage McCarthy, Rachel Nuzum, Stephanie Mika, (Mar. 2007). Sara R. Collins, Karen Davis, and Jennifer Wrenn, and Mary Wakefield. Jennifer L. Kriss. The Building Blocks of Health Reform: Achieving Slowing the Growth of U.S. Health Care Expenditures: Universal Coverage and Health System Savings What Are the Options? ( Jan. 2007). Karen Davis, (May 2008). Karen Davis, Cathy Schoen, and Cathy Schoen, Stuart Guterman, Tony Shih, Sara R. Collins. Stephen C. Schoenbaum, and Ilana Weinbaum. Bending the Curve: Options for Achieving Savings and Why Not the Best? Results from a National Scorecard Improving Value in U.S. Health Spending (Dec. 2007). on U.S. Health System Performance (Sept. 2006). Cathy Schoen, Stuart Guterman, Anthony Shih, The Commonwealth Fund Commission on a High Jennifer Lau, Sophie Kasimow, Anne Gauthier, and Performance Health System. Karen Davis. Framework for a High Performance Health System for A High Performance Health System for the United the United States (Aug. 2006). The Commonwealth States: An Ambitious Agenda for the Next President Fund Commission on a High Performance (Nov. 2007). The Commonwealth Fund Health System. Commission on a High Performance Health System. Public Views on Shaping the Future of the U.S. A Roadmap to Health Insurance for All: Principles for Health System (Aug. 2006). Cathy Schoen, Sabrina Reform (Oct. 2007). Sara R. Collins, Cathy Schoen, K. H. How, Ilana Weinbaum, John E. Craig, Jr., Karen Davis, Anne Gauthier, and Stephen C. and Karen Davis. Schoenbaum. Gaps in Health Insurance: An All-American An Analysis of Leading Congressional Health Care Problem—Findings from the Commonwealth Fund Bills, 2005–2007: Part II, Quality and Efficiency Biennial Health Insurance Survey (Apr. 2006). ( July 2007). Karen Davis, Sara R. Collins, and Sara R. Collins, Karen Davis, Michelle M. Doty, Jennifer L. Kriss. Jennifer L. Kriss, and Alyssa L. Holmgren. Denver Health: A High-Performance Public Health Care System ( July 2007). Rachel Nuzum, Douglas McCarthy, Anne Gauthier, and Christina Beck. 60 ONE EAS T 7 5 T H S T REE T NEW YOR K , NY 1 0 0 2 1 - 2 6 9 2 T EL 2 1 2 . 6 0 6 . 3 8 0 0 FA X 212 . 6 0 6 . 35 0 0 w w w. c o m m o nwe a l t h f u n d .o rg