Can we assure quality?

Transcription

1 Definition

2 Can we assure quality? Standards change Different standards for different locations and same location by different groups Practice is, by its nature, imperfect Factors other than professional inputs are out of the healthcare system s s control, e.g., patient compliance and environmental issues But we can assure the process is under control

3 Patient Care Quality: The degree to which patient care services increase the probability of desired patient outcomes and reduce the probability of undesired outcomes, given the current state of knowledge.

4 A Proposed Definition for Quality in Healthcare Quality is the provision of efficient and effective care to appropriately selected patients at the right time and in an expert manner, consistent with the current state of medical knowledge and patient preferences. J. Shalowitz, MD, MBA

5 Institute of Medicine Criteria- Healthcare should be: Safe Effective Patient-centered Timely Equitable Efficient

6 Safe

7 July 21, 2006 Report Finds a Heavy Toll From Medication Errors By GARDINER HARRIS WASHINGTON, July 20 Medication errors harm 1.5 million people and kill several thousand each year in the United States, costing the nation at least $3.5 billion annually, the Institute of Medicine concluded in a report released on Thursday. Drug errors are so widespread that hospital patients should expect to suffer one every day they remain hospitalized, although error rates vary by hospital and most do not lead to injury, the report concluded.

8 October 31, 2006 What Pilots Can Teach Hospitals About Patient Safety By KATE MURPHY A growing number of health care providers are trying to learn n from aviation accidents and, more specifically, from what the airlines have done to prevent them. In the last five years, several major hospitals have hired professional pilots to train their critical-care care staff members on how to apply aviation safety principles to their work. They learn standard cockpit procedures like communication protocols, checklists and crew briefings to improve patient care, if not save patients lives. Though health care experts disagree on how to incorporate aviation- based safety measures, few argue about the parallels between the two industries or the value of borrowing the best practices. Spurred by a 1999 report by the Institute of Medicine, an arm m of the National Academies, titled To Err Is Human, which estimated that as many as 98,000 patients die annually from preventable medical errors, and by more recent bad publicity from mistakes like amputations of the wrong limbs, many health care providers are redoubling their efforts to improve patient safety.

9 Effective

10 How do we evaluate effectiveness? Compliance with standards. But there are some problems with this suggestion Here are 6 of them:

11 1.Which (whose) standards do we use? (Some examples): National Guideline Clearinghouse ( Zentralstelle der Deutchen Artzeschraft zur Qualitatssicherung in der Medizin, GbR ( Agency for Healthcare Research and Quality-AHRQ ( The Cochrane Collaboration ( National Institute for Clinical Excellence-NICE ( Scottish Intercollegiate Guidelines Network-SIGN ( Canadian Task Force on Preventive Health Care/ Groupe D Etude D Canadien Sur Les Soins De Sante Preventifs ( National Committee for Quality Assurance-NCQA ( New Zealand Guidelines Group ( ) Specialty Societies (for some U.S. Canadian and U.K. sites, see: ) Locally developed, e.g., by physician groups, hospitals/health systems, health plans or government

12 Some other examples: National Quality Forum (NQF) Leapfrog Group: CPOE, High Volume, Intensivists ( AHRQ (formerly AHCPR) ( NCQA: Accreditation of MCOs and HEDIS ( National Patient Safety Foundation ( FACCT (absorbed into the Markle Foundation): Prevention, Staying Healthy, Treating Acute Illness (Getting Better), Living with Chronic Illness, End of Life Issues ( Australian Patient Safety Foundation ( Consumer Assessment of Healthcare Providers and Systems (CAHPS) (

13

14

15 2. What definition do we use?

16

17 WHERE MEDICATION ERRORS OCCUR 13% 1% 25% 61% Ordering Monitoring Administration Dispensing Source: Nebeker et al., Archives of Internal Medicine, May 23, 2005

18 3.How Often Do We Re-Evaluate Standards?

19 Examples: Estrogen Use in Post Menopausal Women Anti-Dysrhythmic Drugs in Asymptomatic Patients Bone Marrow Transplantation Treatment for Metastatic Breast Cancer Bleeding, purging and blistering

20 4. Can the Standards Be Accurately Audited?

21 Example: Questions about data quality forced California to conduct a special study of data accuracy, which found striking variations across hospitals in the validity and reliability of coding certain risk factors overcoding (coding conditions not supported by medical record documentation) rates ranged from 10% at a putatively high-mortality hospital to 74% at a facility considered low mortality. Source: Iezzoni, LI: JAMA:278: , 1607, 1997

22 Importance of Information systems

23 5. Does publishing compliance (not just P4P) change behavior or affect outcomes? Can we achieve the same results without the extra cost?

24 Consumers and purchasers rarely search out the information and do not understand or trust it; it has a small, although increasing, impact on their decision making. Physicians are skeptical about such data and only a small proportion makes use of it. Hospitals appear to be most responsive to the data. In a limited number of studies, the publication of performance data has been associated with an improvement in health outcomes. Source: Marshall, MN et al: The Public Release of Performance Data. JAMA 283: , 1874, 2000

25 The Predictive Accuracy Of The New York State Coronary Artery Bypass Surgery Report-Card System Ashish K. Jha and Arnold M. Epstein Abstract We examined the impact of New York State s s public reporting system for coronary artery bypass surgery fifteen years after its launch. We found that users who picked a top-performing performing hospital or surgeon from the latest available report had approximately half the chance of dying as did those who picked a hospital or surgeon from the bottom quartile. Nevertheless, performance was not associated with a subsequent change in market share. Surgeons with the highest mortality rates were much more likely than other surgeons to retire or leave practice after the release of each report card. Health Affairs,, 25(3): , 855, 2006

26 6. Is the Behavioral Change Sustainable?

27 In the case of physicians, remove the stimulus and the response often goes away.

28 Once we decide which standards to use, we would like to evaluate quality Structure Process Outcome based on:

29 When we evaluate quality standards based on outcomes, there are some problems: 1. Account for factors other than medical care

30 Pay-for-Performance Programs in Family Practices in the United Kingdom (NEJM 355:375-84, 2006)

31 Once we decide which standards to use, we evaluate effectiveness primarily using outcome measures. Here are some problems evaluating them: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention

32 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness Examples: Acute Physiology and Chronic Health Evaluation (APACHE) and Medis Groups: Individual vs. Institutional Evaluations

33

34

35

36 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness 4. Control for vigilance effect 5. Decide what outcome is measured

37 Types of Outcome Measures Functional Status Psychological Status Complications (Morbidity) Death (Mortality) Patient/Family Judgments Appropriate/Efficient Use of Services Source: James Roberts, M.D.

38 Sample SF36 Questions

39 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness 4. Control for vigilance effect 5. Decide what outcome is measured 6. Decide when to measure outcomes- What are the problems with too short or too long intervals?

40 Coated Stents Gain Ground In Risk Trials By KEITH J. WINSTEIN February 13, 2007; Page D4 Coated-stent patients did have more "very late" clots, those that form more than a year after implantation. But this was balanced by a higher incidence of earlier blood clots in bare-metal stents.

41 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness 4. Control for vigilance effect 5. Decide what outcome is measured 6. Decide when to measure outcomes- What are the problems with too short or too long intervals? 7. Decide who chooses the outcome/who is the customer

42 Cancer Patients Gain Say in Drug Approvals FDA and Drug Makers Add Reports From Trial Participants To Traditional Measures Such as Survival, Tumor Shrinkage By AMY DOCKSER MARCUS February 13, 2007; Page D1 A cancer drug's effectiveness has long been measured in two important ways: whether it shrinks a tumor and whether it extends patients' lives. But researchers and regulators are paying increasing attention to another criterion: how a patient feels while taking the medicine. In an important change, cancer patients' own assessments of how a drug is working, called patient-reported outcomes or PROs, are increasingly part of the drug-approval process at the Food and Drug Administration. The agency says PROs have been integral for the approval of a number of cancer drugs in recent years, including Amgen Inc.'s Kepivance for severe mucositis, mouth sores that are a side effect of cancer treatment, and Axcan Scandipharm Inc.'s Photofrin, an agent used in treating precancerous lesions in Barrett's esophagus.

43 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness 4. Control for vigilance effect 5. Decide what outcome is measured 6. Decide when to measure outcomes- What are the problems with too short or too long intervals? 7. Decide who chooses the outcome/who is the customer 8. Choose who or what is being evaluated, e.g., physician (if so which?), institution, payer, system, country

44 Who or What is being evaluated? Example: Pennsylvania Health Care Cost Containment Council (PHC4) Outcomes by Hospital or Physician: Results: -The state s busiest surgeon performed 352 open-heart and CABG procedures in 2003 while three physicians did just one apiece. The average was about 130 procedures per surgeon. -Patients treated by surgeons who performed higher numbers of procedures-in the 200 to 250 range- were twice as likely to survive surgery as those whose physicians performed fewer than 100 procedures that year. Patients with higher-volume Surgeons also tended to have shorter lengths of stay in the hospital. -The statewide average length of stay was 5.9 days, but averages at individual hospital s varied widely from three-point five days to eight days. -The data showed no apparent relationship between higher costs and improved outcomes. Source for summary: Modern Healthcare April 4, 2005

45 Some problems evaluating outcome measures: 1. Account for factors other than medical care 2. Know the natural course of the condition before evaluating outcomes of the intervention 3. Adjust for severity of illness 4. Control for vigilance effect 5. Decide what outcome is measured 6. Decide when to measure outcomes- What are the problems with too short or too long intervals? 7. Decide who chooses the outcome/who is the customer 8. Choose who or what is being evaluated, e.g., physician (if so which?), institution, payer, system, country 9. Choose the threshold for an outcome to register 10. Assess whether it is worth the cost

46 Examples: In 1990, California decided not to implement a risk-adjustment program when the cost was projected to be $61.2 Million. For physicians, is the reward more than the cost of gathering the data?

47 Patient Centered

48 Not all needs are the same The Basics: Access, coordination and thoroughness of care, communication with providers, level of hassles with claims Staying Healthy: Reduction of health risks, early detection and monitoring of illness, avoiding preventable health problems, overall health status Getting Better: Appropriate treatment and follow up, recovery from illness, Living with Illness: Functional status, quality of life Changing Needs: End-of of-life, disabilities, Source: Foundation for Accountability, July 1997

49 Efficient

50 Pay For Performance (P4P)

51 And today 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% U.K. New Zealand Australia Netherlands Germany Canada U.S. Source: 2006 Commonwealth Fund survey of 6000 primary care physicians.

52 How we can improve the effectiveness of P4P: Must align incentives throughout the healthcare system. For example, physician and hospital incentive must be coordinated.

53 Physician Voluntary Reporting Program (PVRP) 16 Measure Core Starter Set Effective April 1, 2006 Note: These measures have been excerpted from the Full 36 Measure Set Medicine Aspirin at arrival for acute myocardial infarction Beta blocker at time of arrival for acute myocardial infarction Hemoglobin A1c control in patient with Type I or Type II diabetes s mellitus Low-density lipoprotein control in patient with Type I or Type II diabetes mellitus High blood pressure control in patient with Type I or Type II diabetes mellitus Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker therapy for left ventricular systolic dysfunction Beta-blocker therapy for patient with prior myocardial infarction Assessment of elderly patients for falls Nephrology (ESRD Program) Dialysis dose in end stage renal disease patient Hematocrit level in end stage renal disease patient Receipt of autogenous arteriovenous fistula in end-stage renal disease patient requiring hemodialysis Psychiatry/Medicine Antidepressant medication during acute phase for patient diagnosed with new episode of major depression Surgery Antibiotic prophylaxis in surgical patient Thromboembolism prophylaxis in surgical patient Use of internal mammary artery in coronary artery bypass graft surgerys Pre-operative beta-blocker blocker for patient with isolated coronary artery bypass graft Source: ationsandinstruction.pdf

54

55 How we can improve the effectiveness of P4P: Pay for performance/align incentives Transparency

56 Percent of Heart Attack Patients Given Aspirin at Arrival 98% 97% 96% 95% 94% 93% 92% 91% 90% 89% Source: AVERAGE FOR ALL REPORTING HOSPITALS IN THE UNITED STATES AVERAGE FOR ALL REPORTING HOSPITALS IN THE STATE OF ILLINOIS EVANSTON NORTHWESTERN HEALTHCARE RUSH NORTH SHORE MEDICAL CENTER

57 September 24, 2006 Give doctor a Web scan Want to do a background check on your doctor? The Minnesota Board of Medical Practice just made it easier for consumers to see profiles of the state's 18,000 physicians and 1,000 physician assistants on its website.

58 How we can improve the effectiveness of P4P: Pay for performance/align incentives Transparency Continuous improvement

59 Transition from: Quality Assurance Quality Improvement Externally Driven Internally Driven Follows Organizational Structure Follows Patient Care Focused on Individuals Focused on Process Delegated to a Few Embraced by All Works toward Endpoints Has No Endpoints Assures Quality (Perfection) Improves Quality Divides Analysis of Integrates Analysis Effectiveness/Efficiency Source: James Roberts, M.D.

60 How we can improve the effectiveness of P4P: Pay for performance/align incentives Transparency Continuous improvement Involve those affected by the process

61 How we can improve the effectiveness of P4P: Align incentives Transparency Continuous improvement Involve those affected by the process Accountability Make information actionable

62 How we can improve the effectiveness of P4P: Align incentives Transparency Continuous improvement Involve those affected by the process Accountability Make information actionable Insure data accuracy and currency Build in incentives to care for sicker patients Reduce variation in care

63 Source: tlas.pdf

64 Source: tlas.pdf

65 Cholesterol high in those at risk for heart ills Wed Feb 8, :20 PM GMT NEW YORK (Reuters Health) - People who are at highest risk for cardiovascular disease generally have the lowest level of control of high cholesterol levels, investigators report. "Given the significance of cardiovascular disease as a public health problem in the US and the proven benefits of lipid-lowering therapy for primary prevention," the researchers comment, "efforts to improve the treatment and control of (high cholesterol) and to eliminate disparities... should be considered among our highest national healthcare quality improvement priorities." Dr. David C. Goff, from Wake Forest University School of Medicine in Winston-Salem, North Carolina, and colleagues evaluated 6704 subjects aged 45 to 84 years who were free of clinical cardiovascular disease at the start of the study between 2000 and Overall, 29 percent of the participants had poor lipid profiles. Of these, only 54 percent were taking lipid-lowering drugs, and of those receiving treatment, only 41 percent achieved their target levels, the researchers report in the American Heart Association's journal Circulation. Poor lipid levels were seen in 12 percent of subjects at low risk for cardiovascular disease, 34 percent of those at intermediate risk, and 49 percent of participants at high risk. More than 80 percent of those in the low-risk group were being treated, compared with only about half of the higher risk groups. Goff and his associates report that this pattern was similar among ethnic groups, except for Chinese Americans, who were less likely to be affected. However, African Americans and Hispanic Americans were less likely to be treated and controlled than non-hispanic whites. This, the team suggests, is likely due to socioeconomic characteristics and healthcare access.

66 Volume-Quality Relationships

67

68 Volume-Quality Relationships What are the possible reasons for volume- quality relationships? What are the public policy implications? Do volume-quality relationships apply to all conditions?

69

70 Volume-Quality Relationships What are the possible reasons for volume- quality relationships? What are the public policy implications? Do volume-quality relationships apply to all conditions? What is more important: physician, hospital and/or both? What about adjustments for patient characteristics other than severity of illness?

71 And finally How do we combine cost, quality and access considerations to assess value as the customer defines it? Competing sample definitions: Given a certain cost, how can I maximize quality? Given a desired quality, how can I minimize cost? What comparable questions would you ask that incorporate access?