CER Module ACCESS TO CARE January 14, 2014. 830 AM 12:30 PM Topics 1. Definition, Model & equity of Access Ron Andersen (8:30 10:30) 2. Effectiveness, Efficiency & future of Access Martin Shapiro (10:30 12:30) 1
Learning Objectives To understand how contextual and individual characteristics influence health behaviors and outcomes of care. To learn about the behavioral model as a conceptual framework for analyzing access and equity in health care. To determine how the dimensions of access can be used to improve utilization and health outcomes. To describe how access has changed over time using nationally standardized indicators that measure potential access (health insurance coverage and regular source of care), realized access (use of hospital, physician, and dental services), and equitable access (providing services according to need for all income and racial and ethnic groups). To assess the effectiveness and efficiency of access in the US To review how the 2010 Affordable Care Act might improve health care access. 2
A definition of access to care We define access as the actual use of personal health services and everything that facilitates or impedes the use of personal health services. It is the link between health services systems and the populations they serve. The conceptualization and measurement of access is key to the understanding and formulating health policy because it can be used to predict health services use, to promote social justice and to improve health outcomes. 3
Figure 1 The Behavioral Model of Access to Care 6 th Revision 2013 Contextual Characteristics Individual Characteristics Health Behavior Outcomes PREDISPOSING ENABLING NEED Demographic Health Policy Environmental Social Financing Population Health Indices Beliefs Organization PREDISPOSING ENABLING NEED Demographic Financing Perceived Social Organization Evaluated Beliefs Genetics Personal Health Practices Process of Medical Care Use of Personal Health Services \ Perceived Health Evaluated Health Consumer Satisfaction Quality of Life 4
Model Contextual and Individual Characteristics The Model stresses that improving access to care is best accomplished by focusing on contextual as well as individual determinants Context includes health organization and provider-related factors as well as community characteristics Contextual characteristics are divided in the same way as individual characteristics: (1) existing conditions that predispose people to use or not use services; (2) enabling conditions that facilitate or impede use of services, and (3) need or conditions that laypeople or health care providers recognize as requiring medical treatment (Andersen, 1968; Andersen, 1995). 5
Model Contextual Characteristics Predisposing Enabling Need Arrows leading from contextual characteristics 6
Model Individual Characteristics Predisposing Enabling Need Arrows leading from individual characteristics 7
Model Health Behaviors Personal health practices Process of medical care Use of personal health services 8
Model Outcomes Perceived health Evaluated health Consumer satisfaction Quality of life Arrows from outcomes 9
Dimensions of Access: Measurement, Trends, and Strategies for Improvement Defining measuring and improving dimensions of access (Andersen) Potential and realized access: trends and equity (Andersen) Effective and efficient access: findings and equity (Shapiro) 10
Dimensions of Access: Defining, Measuring & Improving 1. Dimension Potential Access (Enabling Factors) Intended Improvement To Increase or Decrease Health Services Use 2. Realized Access To Monitor and Evaluate Policies to (Use of Services) Influence Health Services Use 3. Equitable Access To Insure Health Services Distribution is is Determined by Need 4. Inequitable Access To Reduce the Influence of Social Characteristics and Enabling Resources on Health Services Distribution 5. Effective Access To improve the Outcomes (Health Status, Satisfaction, Quality of Life ) from Health Services Use 6. Efficient Access To Minimize the Costs of Improving Outcomes from Health Services Use 11
Potential and realized access trends and equity Potential access trends (Table 1) Realized access trends (Tables 2, 3) Potential access equity (Table 1) Realized access equity (Tables 2,3,4) 12
Table 2.1. Percent of the U.S. population under sixty-five with no health insurance coverage and regular source of medical care No Insurance No Regular Source of Care 1984 2002 2009 2011 1993-1994 1999-2000 2010 Age <18 years 14% 13% 10% 7% 7.7% 6.9% 9.9% 18 44 years 17 22 24 25 21.7 21.4 25.2 45 64 years 10 13 13 15 12.8 10.9 11.8 Race and ethnicity d White, non- Hispanic 12 12 12 13 17.1 14.9 15.8 Black, non- Hispanic 20 19 18 19 19.7 19.2 22.1 Hispanic, Mexican b 34 40 38 33 Hispanic, Puerto Rican b Hispanic, Cuban b 18 16 21 16 } 30.3 32.6 33.3 22 25 23 26 Asian a 18 18 17 17 24.8 22.1 20.8 American Indian/ Alaskan Native a 22 38 35 34 - - - Percent of poverty level c, d Below 100 34 34 32 28 29.5 29.6 32.8 100 199 22 31 29 30 25.4 27.1 30.4 200 299 8 15 16 17 15.6 17.2 19.3 300 and more 3 6 6 5 13.4 11.6 9.7 Total 14 17 17 17 17.9 17.8 19.6 13
Source: National Center for Health Statistics, Health United States, 2011 (Hyattsville, MD: National Center for Health Statistics, 2011). Updated tables 138, 140, 141. Note: Table 2.1 reports a critical potential access measure: health care coverage for persons under sixty-five years of age from 1984 to 2011. The uninsured proportion of the population increased from 14 to 17 percent in that time period. Medicaid coverage actually increased (from 7 to 18 percent), but the overall decline in coverage resulted from a drop in the proportion covered by private insurance, from 77 to 62 percent (data not shown). a Includes persons of Hispanic and non-hispanic origin. b Persons of Hispanic origin may be white, black, Asian, Pacific Islander, or American Indian/Alaskan Native. c Poverty level is based on family income and family size, using Bureau of the Census poverty thresholds. d Age-adjusted by U.S. Census 2000 data. The proportion of population eighteen to forty-four years who were uninsured increased during the 1980s and 1990s, reaching 25 percent in 2011 (Table 2.1). The proportion covered by private insurance decreased for every age group between 1984 and 2011 (data not shown). Between 1984 and 2011, the proportion of all children 14
Table 2.2. One Percent or of More the U.S. Admissions population with hospital admissions, Two or More physician Admissions visits, and dental visits by poverty 1997 2011 1997 2011 Acute Hospital Admissions in a Year b Percent of poverty level a <100 10.0% 8.5% 2.8% 2.9% 100 199 7.2 6.6 1.7 1.8 200 399 6.0 5.3 1.2 1.2 400 or more 4.7 4.4 0.7 0.8 Total 7.8 7.1 1.8 1.9 Physician Visits in a Year b No Visits Ten or More Visits Percent of poverty level a <100 21% 19% 20% 17% 100 199 20 21 15 14 200 399 17 16 13 12 400 or more 13 11 13 12 Total 17 16 14 13 Dental Visits in a Year b Percent of poverty level a No Visits One or More visits 15
100 199 49 48 51 52 200 399 34 35 66 65 400 or more 21 19 79 81 Total 35 34 65 66 Source: National Center for Health Statistics, Health United States, 2011 (Hyattsville, MD: National Center for Health Statistics, 2011), Updated tables 102, 83, 98. a Poverty level is based on family income and family size, using the census poverty thresholds. b Age-adjusted by U.S. Census 2000 data. 16
Table 2.3. Percent of the One U.S. or population More with hospital admissions, Two or physician More visits, and dental visits by race o 1997 2011 1997 2011 Acute Hospital Admissions in a Year c Race/Ethnicity White, non-hispanic 6.1% 5.8% 1.2% 1.3% Black, non-hispanic 7.5 6.8 1.9 2.2 Hispanic b 6.8 4.9 1.2 1.2 Asian a 3.9 3.4 0.5 0.6 American Indian/Alaskan Native a 7.6 4.9 - - Total 7.8 9.3 1.3 2.0 Physician Visits in a Year c None Ten or More Race/Ethnicity White, non-hispanic 15% 13% 14% 14% Black, non-hispanic 17 15 14 13 Hispanic b 25 23 13 10 Asian a 23 20 9 8 American Indian/Alaskan Native a 17 19 21 12 Total 17 16 14 13 Dental Visits in a Year c None One or More Race/Ethnicity White non-hispanic 32% 31% 68% 69% Black non-hispanic 41 38 59 62 Hispanic b 46 43 54 57 17
Asian a 37 34 63 66 American Indian/Alaskan Native a 45 38 55 62 Total 35 34 65 66 Source: National Center for Health Statistics, Health United States, 2011 (Hyattsville, MD: National Center for Health Statistics, 2011). Updated tables 102, 83, 98. a Includes persons of Hispanic and non-hispanic origin. b Persons of Hispanic origin may be white, black, Asian, Pacific Islander, or American Indian/Alaskan Native. c Age-adjusted by U.S. Census data 2000 18
Table 2.4. Selected measures of need by race/ethnicity, and poverty level Infant Death per 1,000 Births per Year Age- Standardize Fair or Poor Self- Any Basic Action Untreated Dental Any Joint Pain d Death Rate Reported Difficulty or Caries (18 and per Year [SB1]Make Health sure the stars Complex are in for the other (20 64) hispanic groups Over) Activity Limitation Serious Psychologica l Distress in 30 Days 2006 2008 2007 2009 2011 2011 2005 2008 2010 2009 10 Race or Ethnicity White, non-hispanic 5.6 7.6 8.4% a 32.4% a 19.3% 32.6% 3.1% Black, non-hispanic 13.1 9.4 15.0 33.3 39.7 32.0 3.8 Hispanic, Mexican b * * 14.0 35.2 25.0 2.8 Hispanic, Puerto } Rican b 5.5 5.3 13.2 24.4 * 25.4 3.6 Hispanic, Cuban b * * * * * * Asian a 4.6 4.1 8.7 19.0 * 20.4 1.6 American Indian/Alaskan Native a 8.6 6.1 14.4 40.8 * 38.1 5.2 Percent of Poverty Level c Below 100 * * 21.5% 41.0% 41.9% 35.6% 8.4% 100 199 * * 15.0 40.0 37.7 34.0 4.8 200 399 * * 8.7 31.4 24.3 32.2 2.8 400 and up * * 4.3 21.7 24.3 30.5 1.2 Total * * 9.8 31.9 11.1 32.1 3.2 19
Source: National Center for Health Statistics, Health United States, 2011 (Hyattsville, MD: National Center for Health Statistics, 2011), Updated tables 15, 24, 25, 56, 54, 57, 76, 53, 59. a Includes persons of Hispanic and non- Hispanic origin. b Persons of Hispanic origin may be white, black, Asian, Pacific Islander, or American Indian/Alaskan Native. c Poverty level is based on family income and family size, using Bureau of the Census poverty thresholds. * Not reported. 20
Effective and efficient access: measures and equity (Shapiro) Effective access measures Effective access equity (Table 5, cols. 1-3) Efficient access measures Efficient asccess equity (Table 5 cols. 4-5) 21
Effective access measures the IOM Committee on Monitoring Access to Medical Care defined effective access as the timely use of personal health services to achieve the best possible health outcomes (Committee on Monitoring Access to Personal Health Care Services & Institute of Medicine, 1993). Measures of effective access examine the effect of potential access (health insurance and regular source of care) and realized access (health services utilization) on outcomes (health status, quality of life, and patient satisfaction with health services). 22
Effective access equity (cols. 1-3) Table 2.5. Selected measures of effectiveness and efficiency by race/ethnicity and incom Cancer (Age 50 and over Who Received Colonoscopy, Sigmoidoscopy, or Proctoscopy) Effectiveness Cardiovascula r Disease Deaths per 1,000 Hospital Admissions for Acute Myocardial Infarction (18 and over) Diabetes (Age 40 and over with Diagnosed Diabetes Who Received All Four Recommend ed Services in a Calendar Year) Potentially Avoidable Hospitalizati ons per 1,000 Population (18 and over) Efficiency Adults Aged 65 and over with at Least One Prescription from 33 Medications Potentially Inappropriate for Older Adults 2008 2008 2008 2008 2008 Race and Ethnicity White, non- Hispanic 58.4% 59.9% 22.7% 12.6% 13.5% Black, non- Hispanic 50.4 48.4 16.6 25.7 15.6 Hispanic b 38.8 59.0 18.7 13.8 15.4 Asian a 48.1 61.1 * 13.8 8.3 American Indian/Alaskan Native a 35.3 * * * * Percent of Poverty Level c Below 100 36.7 * 16.2 * 16.8 100 199 45.5 * 17.9 * 14..8 200 399 53.3 * 17.0 * 14.4 400 and up 65.1 * 27.8 * 11.6 Median Income of Patient Zip Code First quartile * 61.6 * 20.5 * Second quartile * 60.1 * 13.9 * Third quartile * 56.1 * 12.0 * Fourth quartile * 56.2 * 10.7 * Total 55.4 58.7 21.0 14.2 13.7 23
Source: AHRQ, National Health Care Disparity Report 2011 (AHRQ Publication No. 12-0006). 2012. Retrieved from http://www.ahrq.gov/research/findings/nh qrdr/nhqrdr11/. Appendix data Tables 1_3_2, 2_2_2.2, 4_1_1, 17_2_1, 17_2_2, 15_1_2.[AU: The underscoring in this source note looks incorrect; please revise as needed.] These are correct a Includes persons of Hispanic and non- Hispanic origin. b Persons of Hispanic origin may be white, black, Asian, Pacific Islander, or American Indian/Alaskan Native. c Poverty level is based on family income and family size, using Bureau of the Census poverty thresholds. 24
Percentage of patients who receive recommended (effective) care Care received Recommended care not received Female Male White Black Hispanic Other <$15,000 $15,000-$50,000 >$50,000 Uninsured Medicaid Medicare 56.6% 52.3% 54.1% 57.6% 57.5% 55.4% 53.1% 54.7% 56.6% 53.7% 54.9% 56.9% 43.4% 47.7% 45.9% 42.4% 42.5% 44.6% 46.9% 45.3% 43.4% 46.3% 45.1% 43.1% Overall, people receive 55% of recommended care. All demographic groups, not just minorities or the poor, are at risk for poor quality health care. The differences among sociodemographic subgroups in quality of health care are small in comparison to the 42% - 45% gap between what everyone needs and the care everyone is receiving. Managed care 55.2% 44.8% Private non-managed care 53.6% 46.6% 0% 20% 40% 60% 80% 100% Source: McGlynn, et al., Who Is at Greatest Risk for Receiving Poor-Quality Health Care? The New England Journal of Medicine March 16, 2006 25 (354:11) 25
Risk adjusted rates of adverse events/complications (receiving ineffective care) per 10,000 patients 600 500 400 300 200 100 0 White non-hispanic Black non-hispanic Asian/Pacific Islander non-hispanic Hispanic 555 18 31 Infections due to medical care 96 164 31 43 26 90 Postoperative pulmonary embolus or deep vein thrombosis 241 194 328 Decubitus ulcers (pressure sores) Compared to White elderly patients: Minority patients were more likely to acquire infections in the hospital. Black patients were more likely to suffer blood clots in their legs or lungs following surgery. Black and Hispanic patients were more likely to develop pressure sores. Source: Leatherman and McCarthy, Quality of Health Care for Medicare Beneficiaries: A Chartbook, 2005 The Commonwealth Fund. 26 26
Efficient accesss measures. Aday and colleagues describe efficiency as producing the combination of goods and services with the highest attainable total value, given limited resources and technology (Aday, Begley, Lairson, & Slater, 1998).. Like effective access, efficient access measures the impact of potential access (enabling resources like health insurance) and realized access (health services utilization) on outcomes (health status, quality of life, and satisfaction). The difference is that efficient access values the input (potential or realized access) so that the cost of producing one unit of the outcome can be calculated (such as one person-year of life saved, reduction of one disability day per person per year) 27
Efficient access equity (cols. 4-5) Table 2.5. Selected measures of effectiveness and efficiency by race/ethnicity and incom Cancer (Age 50 and over Who Received Colonoscopy, Sigmoidoscopy, or Proctoscopy) Effectiveness Cardiovascula r Disease Deaths per 1,000 Hospital Admissions for Acute Myocardial Infarction (18 and over) Diabetes (Age 40 and over with Diagnosed Diabetes Who Received All Four Recommend ed Services in a Calendar Year) Potentially Avoidable Hospitalizati ons per 1,000 Population (18 and over) Efficiency Adults Aged 65 and over with at Least One Prescription from 33 Medications Potentially Inappropriate for Older Adults 2008 2008 2008 2008 2008 Race and Ethnicity White, non- Hispanic 58.4% 59.9% 22.7% 12.6% 13.5% Black, non- Hispanic 50.4 48.4 16.6 25.7 15.6 Hispanic b 38.8 59.0 18.7 13.8 15.4 Asian a 48.1 61.1 * 13.8 8.3 American Indian/Alaskan Native a 35.3 * * * * Percent of Poverty Level c Below 100 36.7 * 16.2 * 16.8 100 199 45.5 * 17.9 * 14..8 200 399 53.3 * 17.0 * 14.4 400 and up 65.1 * 27.8 * 11.6 Median Income of Patient Zip Code First quartile * 61.6 * 20.5 * Second quartile * 60.1 * 13.9 * Third quartile * 56.1 * 12.0 * Fourth quartile * 56.2 * 10.7 * Total 55.4 58.7 21.0 14.2 13.7 28
Examples of efficient and inefficient QALY production What is a quality-adjusted life year saved? Is the quality adjustment problemmatic? Does this matter if diff. are large? Richards-Kortum, ch. 5 29
Conclusions and predictions about access to care (1) (Shapiro) Potential access: A key potential access measure, health insurance, reveals that although a growing number of people are being covered by Medicaid, there has been a decline in the number covered by private insurance in the last twenty-five years and an overall increase in the proportion without any health insurance coverage. The Affordable Care Act (ACA) will make health insurance more affordable to populations uninsured or underinsured. Realized access: Although we have documented continuing increases in some realized access measures notably physician and dental visits inpatient hospital admissions and length of stay have been declining for thirty years. However, the declining hospital use rate reflects not so much a reduction in access to appropriate care, but a shift to outpatient services and greater emphasis on primary care and ambulatory outpatient care, reducing the need for acute inpatient services. Equitable access: Low-income and black populations appear to have achieved equity of access according to gross measures of hospital and physician utilization, but continue to lag considerably in receipt of dental care. Hispanics continue to lag in all of these gross measures of access. However, if the generally greater needs of all the minority groups are taken into account, 30 inequities continue to exist for all minorities.
Conclusions and Predictions about access to care (2) (Shapiro) The effectiveness of access is making progress over time. However, comparisons of mortality and morbidity rates between the United States and other developed countries call into question the relative effectiveness of the U.S. health care system. Also, some measurements of the treatments and outcomes for particular diseases suggest inequities in effectiveness for some minority and low-income groups. Implementation of the ACA might lead to improvement in effectiveness, for example through newborn screening recommended by the Uniform Screening Panel to detect severe monogenetic diseases (Stark, 2012). Efficient access: The United States is by far the most costly system in the world, with continually spiraling prices and little evidence that the system is generating the kinds of outcomes that might be expected from such a costly system. There is evidence that some types of health care organization and finance (enabling resources) within the system are more efficient than others, but these forms have not been adopted on a scale to greatly influence the national 31
Some discussion questions on access (Shapiro) 1. Improving access to care of individuals depends more on contextual than on individual determinants. What is the rationale for this statement? Do you agree or disagree? 2. What is meant by the observation Equity of access is in the eyes of the beholder? How equitable is the U.S. health care system according to your definition of equity? 3. A health policy might promote realized access but not effective access or efficient access. The majority of the U.S. population believes the U.S. system has a major problem. Is the most important problem the system faces one of realized access, effective access, or efficient access? 4. As the United States moves increasingly toward evidenced-based medicine, what dimensions of access become increasingly important? What particular measures of access and what health services research studies would you propose to support this movement? 5. What is the good news and what is the bad news when we examine the trends in various measures of access over time? What might enable us to continue the good news and change the bad news to good? 32