People living with chronic conditions are particularly vulnerable

Transcription

1 Rising Out-Of-Pocket For Chronic Conditions: A Ten- Year Trend The prevalence of chronic conditions in the United States has increased since 996, and not just among the oldest old. by Kathryn Anne Paez, Lan Zhao, and Wenke Hwang ABSTRACT: We examined the prevalence of self-reported chronic conditions and out-ofpocket spending using the 5 Medical Expenditure Panel Survey (MEPS) and made comparisons to previously published MEPS data. Our study found that the prevalence of selfreported chronic conditions is increasing among not only the old-old but also people in midlife and earlier old age. The greatest growth occurred in the number of people affected by multiple chronic diseases, a group with sizable out-of-pocket spending. Policymakers should be aware that cost sharing at the point of care can disproportionately burden people with chronic conditions and discourage adherence to drugs that prevent disease progression. [Health Affairs 8, no. (9): 5 5;.377/hlthaff.8..5] People living with chronic conditions are particularly vulnerable to rising out-of-pocket medical spending. Employer-sponsored insurance and Medicare health insurance premiums are generally shared across all policyholders; however, increased out-of-pocket spending at the point of care differentially burdens people with chronic conditions who require multiple services to maintain optimal health and treat disease. In recent years, employers have coped with the resurgence in medical inflation by increasing employees premium contributions and out-of-pocket obligations. Although the dollar amount spent by employees on health insurance premiums has increased, the employee contribution as a percentage of premium has fallen since the 99s. The more dramatic increase in privately insured consumers spending in this decade is from out-ofpocket obligations such as multitier pharmacy plans, a shift from copayments to coinsurance, increased deductibles, combining deductibles with copayments, and smaller provider networks leading consumers to seek out-of-network care. 3 Increasing out-of-pocket spending is intended to discourage unnecessary dis- KathrynPaez(kpaez@s-3.com)andLanZhaoareseniorresearchscientistsintheCenterforHealthPolicyand Research, Social and Scientific Systems, in Silver Spring, Maryland. Wenke Hwang is an assistant professor in the Department of Social Sciences and Health Policy, Division of Public Health Sciences, at the Wake Forest University School of Medicine in Winston-Salem, North Carolina. HEALTH AFFAIRS ~ Volume 8, Number 5 DOI.377/hlthaff Project HOPE The People-to-People Health Foundation, Inc.

2 Scope Of Problem cretionary spending, but it may also reduce the use of clinically important services and drugs that prevent new onset and progression of chronic disease. Most research investigating the effects of increased cost sharing has focused on prescription drug use. Higher drug copayments and three-tier pharmacy plans have been found to reduce adherence to drugs for management of such chronic conditions as diabetes, hypercholesterolemia, hypertension, and schizophrenia. 5 Reduced drug adherence includes delaying prescriptions fills, failing to fill prescriptions, cutting dosages, and reducing the frequency of administration. In an earlier paper by Wenke Hwang and colleagues, out-of-pocket medical spending by people diagnosed with chronic conditions was examined using data from the 996 Medical Expenditure Panel Survey (MEPS). 6 Results showed that out-of-pocket spending, particularly drug costs, increased with the number of chronic diseases and was substantial for both the elderly and the nonelderly. In this paper we present more recent data to reassess the impact of chronic conditions on out-of-pocket spending and to identify trends that should be considered by policymakers and other decisionmakers. Replicating the methodology developed in an earlier study enabled us to compare changes in chronic disease burden and out-of-pocket spending over a ten-year period. Study Data And Methods Data. We used data from the 5 MEPS, sponsored by the Agency for Healthcare Research and Quality (AHRQ). This nationally representative household survey collects detailed information on health status, health insurance coverage, and health care use and spending. An in-depth description of the survey methodology can be found elsewhere. 7 Our analyses are weighted to represent the 9 million civilian noninstitutionalized U.S. population. Definition of chronic conditions. Chronic conditions were defined as conditions that had lasted or [were] expected to last twelve or more months and resulted in functional limitations and/or the need for ongoing medical care. 8 Apanelofphysicians reviewed the three-digit International Classification of Diseases, Ninth Revision (ICD-9) codes and classified codes as meeting the chronic condition definition in adults and 77 in children. AHRQ s Clinical Classification System (CCS) was then used to determine the number of distinct chronic conditions per person. 9 Multiple conditions that fell into the same CCS category were counted as one condition. Peopleweredesignatedashaving,,,or3ormorechronicconditions. Out-of-pocket spending. Out-of-pocket expenditures were self-reported payments for coinsurance, copayments, deductibles, and medically related items and services not covered by insurance. Health insurance premiums were not included, because the focus of this study was to measure the financial burden that is directly related to medical care use. People not using any medical services during the year were excluded from this segment of the study. Extreme data were capped at the 99.5 percentile to reduce the influence of outliers on mean out-of-pocket spending. 6 January/February 9

3 Out-of-pocket spending indices were created to examine increases in spending that take into consideration changes in both reported disease prevalence and outof-pocket spending over time. The 996 mean expenditures taken from Hwang and colleagues out-of-pocket spending study were inflated to 5 U.S. dollars based on the medical component of the Consumer Price Index (CPI). Average mean spending was weighted by disease prevalence for each year, and then ratios comparing 5 spending with 996 spending were calculated. Insurance and poverty status. Respondents were grouped into age categories and then further divided into mutually exclusive insurance categories. Each person needed to have more than six months of an insurance type to be assigned to the private insurance and Medicare insurance groups. The Medicaid coverage criterion was three months, because eligibility can vary from month to month. The classification of uninsured was applied only if no coverage was reported for the entire year. People under age sixty-five were categorized as having private, Medicaid, or other public insurance (Medicare, Tricare, or private and other public/physician programs), or uninsured. Those over age sixty-five were grouped as having Medicare only, both Medicare and private, or both Medicare and Medicaid. Poverty status was classified based on family income and the federal poverty level. Study Results Population with chronic conditions. In5,3.8percentoftheU.S.civilian, noninstitutionalized population had one or more condition that we classified as chronic (Exhibit ). One in five reported living with one chronic condition, while.7 percent of respondents reported two conditions, and 3.3 percent had three or more (multiple) conditions. Among adults, hypertension, hyperlipidemia, and diabetes mellitus without complications were the most prevalent conditions, accounting for 3. percent of all reported chronic conditions. Among those under age twenty, no conditions clearly predominated; upper respiratory disease, asthma, and attention deficit hyperactivity disorder (ADHD) were most common but together accounted for only.7 percent of all conditions. Thepresenceofchronicdiseaseincreasedwithageforthenonelderlypopulation. The most dramatic rise occurred between early adulthood (ages ) and midlife (ages 5 6): an increase from 3. percent to 63. percent of people with at least one chronic condition. After age sixty-four, people were most likely to be burdened by multiple chronic conditions (5.3 percent in younger-old age, ages 65 79; 5. percent in the old-old age group, age eighty and older) and least likely to report no chronic conditions. Dual eligibles (people with both Medicare and Medicaid) older than age sixty-four carried the most disease burden of all insurance groups: 5.7 percent reported having multiple chronic diseases. People under agesixty-fiveanduninsuredtendedtobehealthierthanallothergroups;.7percent of this group reported having a condition classified as chronic. When ethnicity and race were considered, non-hispanics reported greater HEALTH AFFAIRS ~ Volume 8, Number 7

4 Scope Of Problem EXHIBIT Socioeconomic Characteristics Of Medical Expenditure Panel Survey (MEPS) Respondents, By Number Of Chronic Conditions, 5 Number of chronic conditions Characteristic Number (millions) None One Two Three or more Total population % 9.7%.7% 3.3% Age (years) Sex Male Female Race White Black Other Hispanic ethnicity Hispanic Non-Hispanic Insurance status a Age 65 and older Medicare only Medicare/private Medicare/Medicaid Private Medicaid Other public Uninsured Poverty status b Poor Near-poor Low income Middle income High income SOURCE: Authors tabulations of 5 MEPS Household Component survey data. a Excludes people with unknown type of insurance. b As proportion of federal poverty level; see Note in text. chronic condition burden than Hispanics (6.7 percent versus 6.8 percent). Whites reported having more chronic conditions than both blacks (6.5 percent versus 37. percent) and other races (3. percent). Women were more likely than men to report having a chronic condition, and, in particular, multiple conditions (5.8 percent versus.7 percent). The association of ethnicity, race, and sex with chronic conditions remained when age was controlled for in regression models (data not shown). Out-of-pocket spending by individual characteristics. rose as the 8 January/February 9

5 number of chronic conditions increased, with the largest increase occurring between zero and one condition (9 percent increase), and between two and multiple conditions (79 percent increase; Exhibit ). varied by sociodemographic characteristics most notably, increasing linearly with advancing age. Women, whites, and non-hispanics paid more out of pocket for medical services than their counterparts. Mean out-of-pocket spending for other public insured and unin- EXHIBIT Socioeconomic Characteristics And Mean Annual Out-Of-Pocket Per Person, By Number Of Chronic Conditions, 5 Number of chronic conditions Characteristic All None One Two Three or more Total population $ 7 $33 $ 655 $,39 $,865 Age or older 88 55,,79, , ,8,7,36 95,6,77,9,39 Sex Male Female ,3,733,95 Race White Black Other , ,938,39,55 Hispanic ethnicity Hispanic Non-Hispanic ,5,5,89 Insurance status a Age 65 and older Medicare only Medicare/private Medicare/Medicaid,856, b, ,59,53 65,588,7 85 Private Medicaid Other public Uninsured 636 8, ,69, ,57,6,673 87,69,38 Poverty status c Poor Near-poor Low income Middle income High income ,9,7,7,86,,68,86,96, SOURCE: Authors tabulations of 5 Medical Expenditure Panel Survey (MEPS) Household Component survey data. a Excludes people with unknown type of insurance. b Sample size of fewer than people with a standard error greater than 3 percent, which is insufficient for making reliable national estimates. c As proportion of federal poverty level; see Note in text. HEALTH AFFAIRS ~ Volume 8, Number 9

6 Scope Of Problem sured people under age sixty-five with any number of chronic conditions were higher than the costs paid by the privately insured or Medicaid recipients. The other public group was small ( percent of the total population) and heterogeneous. When spending among people with multiple chronic conditions was considered by poverty status, the poor, near-poor, and low-income spent approximately double what Medicaid recipients spent ($87) and were within 7 9 percent of spending by those classified with high incomes ($,). Out-of-pocket spending by type of service. Drugswerethecostliesttypeof medical expenditure for almost all groups (Exhibit 3). People over age sixty-five with multiple chronic conditions spent an annual average of $,9 per person for drugs more than any other group and more than five times greater than their spendingforofficevisits.thenewmedicarepartddrugbenefitmaymitigatethefinancial burden of drug costs for this group. Dental care followed drugs as the second most costly out-of-pocket health care expenditure but is less likely than other categories presented to be directly associated with chronic condition burden. EXHIBIT 3 Percentage Using Services And Mean Out-Of-Pocket Per Person, By Type Of Medical Service And Number Of Chronic Conditions, 5 Percent using services Number of chronic conditions Rx drugs Home health Dental services Office visits Hospital inpatient Hospital outpatient and ED Medical equipment Vision aids Age 65 or older 3 or more 95% 6 93 % 8 5 7% % % 5 9 7% % 7 7 % 7 3 or more Mean out-of-pocket spending Age 65 or older 3 or more $ ,9 $6 9 a a 8 a 8 $ $ $ $ $5 5 a a 3 8 $ or more a SOURCE: Authors tabulations of 5 Medical Expenditure Panel Survey (MEPS) Household Component survey data. NOTES: Data are for the 85 percent of the population who used any type of service. ED is emergency department. a Sample size of fewer than people with a standard error greater than 3 percent, which is insufficient for making reliable national estimates. January/February 9

7 Comparisons between self-reported chronic disease prevalence in 996 and 5. Exhibit presents the percentage-point increase in chronic condition prevalence from 996 to 5 by reported number of chronic conditions. An overall shift occurred from people reporting zero or only one chronic condition to people reporting multiple chronic conditions, particularly among people in midlife and older. Reports of multiple chronic conditions rose 9.7 percentage points among EXHIBIT Socioeconomic Characteristics Of Medical Expenditure Panel Survey (MEPS) Respondents, Differences In The Percentage Of People With Chronic Conditions Between 996 And 5 Number of chronic conditions Characteristic None One Two Three or more Total population 3.%.%.% 5.9% Age (years) or older Sex Male Female Race White Black Other Hispanic ethnicity Hispanic Non-Hispanic Insurance status a Age 65 and older Medicare only Medicare/private Medicare/Medicaid Private Medicaid Other public Uninsured Poverty status b Poor Near-poor Low income Middle income High income SOURCE: Authors tabulations of 5 Medical Expenditure Panel Survey (MEPS) Household Component survey data. Data for 996 are from W. Hwang et al., Out-of-Pocket Medical for Care of Chronic Conditions, Health Affairs, no. 6 (): a Excludes people with unknown type of insurance. b As proportion of federal poverty level; see Note in text. HEALTH AFFAIRS ~ Volume 8, Number

8 Scope Of Problem those in midlife; 7.6 percentage points among the younger-old; and 6.6 percentage points among the old-old. A dramatic upward shift toward people reporting multiple chronic conditions was also found among Medicare beneficiaries. The increase in three or more chronic conditions occurred nearly equally across sex, race, ethnicity, and income groupings. Trends in out-of-pocket spending, People using health services spent an average of $7 per person in 5 for health care services (Exhibit ), comparedto$7in After adjusting for inflation, this represents a 39. percent increase in out-of-pocket spending per person, which is consistent with a 35.3 percent rise in the National Health Expenditure Accounts for the same ten-year time period. An out-of-pocket expenditure index (EI) was created to measure the overall increase in out-of-pocket spending comparing 5 to 996 spending, holding disease prevalence constant (Exhibit 5). 5 Thechangeseenisduesolelytoincreasing out-of-pocket spending. The EI for the overall population was.9, indicating that expenditures were 9 percent higher in 5 than in 996, when chronic condition prevalence was held constant. The younger-old had the greatest increase, with an EI of.3, followed by young adults, those in midlife, and the old-old. All insurance categories, including Medicaid recipients, had a sizable increase in out-of-pocket spending over the ten-year period. The largest increase was experienced by those in the other public insurance category, followed by the uninsured and Medicare-only beneficiaries. Although smaller, the EI for Medicaid recipients was substantial when controlling for rising chronic condition prevalence. When income was considered, the EI was highest for people classified as poor the group least able to absorb increased medical costs. Low-income people had the second-highest EI. Although the EI for the near-poor group was the lowest of all income groups, it still reflected a marked increase in out-of-pocket spending. An EI was created for drugs since out-of-pocket spending was greatest for drugs of all expenditure types and was particularly high for people with multiple chronic diseases. for drugs went up considerably for both people under agesixty-fiveandpeopleagesixty-fiveandolder. Summary And Policy Implications Rising prevalence and spending. Our study found that out-of-pocket spending and chronic disease prevalence are increasing among not only the old-old but among people in midlife and early old age, without regard to sex, race, ethnicity, or income. The greatest growth occurred in the number of people reporting multiple chronic diseases; this is also the group with the most substantial out-of-pocket spending. Overall, out-of-pocket spending increased by 39. percent per person over the ten-year period. The growth in out-of-pocket spending was not evenly distributed across the population. increases were 9 percent higher overall when holding the rising prevalence of chronic conditions constant, with the great- January/February 9

9 EXHIBIT 5 Out-of-Pocket Expenditure Index (EI) By Population Characteristics, Comparing 5 To 996 Characteristic Index Total population.9 Age (years) or older Insurance status Age 65 and older Medicare only Medicare/private Medicare/Medicaid Private Medicaid Other public Uninsured Poverty status a Poor Near-poor Low income Middle income High income Services Rx drugs Age 65 and older SOURCE: Authors tabulations of 5 Medical Expenditure Panel Survey (MEPS) Household Component survey data. Data for 996 are from W. Hwang et al., Out-of-Pocket Medical for Care of Chronic Conditions, Health Affairs, no. 6 (): NOTES: 996 is the base year. 996 dollars were inflated to 5 dollars based on the medical component of the Consumer Price Index. a As proportion of federal poverty level; see Note in text. est increases among those in early old age, the other public insured, the uninsured, Medicare beneficiaries, the poor, and people who take prescription drugs. Medicaid continued to provide financial protection for people with chronic conditions from high out-of-pocket spending. When poverty status was considered, it became evident that Medicaid is not available to all poor people with chronic conditions. Need for coverage expansion and redesign. Our findings highlight the need to expand coverage to nonelderly adults who are unable to obtain health insurance through employers or other means. This age group is increasingly developing chronic conditions while becoming more likely to be uninsured. 6 The nonelderly uninsured frequently fail to get needed medical care and drugs for chronic health HEALTH AFFAIRS ~ Volume 8, Number 3

10 Scope Of Problem conditions because of cost. 7 A proper exploration of this issue is beyond the scope of this paper. Our findings also suggest that efforts to make health care affordable must be strategic and judicious. Benefit redesign should include broadening coverage for appropriate chronic care to address the rising prevalence of chronic conditions among adults beginning in middle age. Efforts to control rising health spending by increasing out-of-pocket spending for essential services and medications may have unintended long-term consequences; raising copayments and deductibles for drugs and services reduces adherence and can lead to poor disease control. 8 Insurers should consider value-based insurance designs that subsidize high-value chronic care while increasing cost sharing for elective services without proven benefit. 9 Study limitations. Several limitations should be considered in interpreting our findings. First, chronic conditions were self-reported, introducing potential reporting bias. Some groups, such as racial/ethnic minorities, the less educated, people with lower incomes, and the uninsured, might hesitate to report chronic conditions or might not realize that they have a chronic condition. In MEPS, reporting bias was minimized by contacting a sample of providers to populate missing data or validate self-reported data, or both. Second, out-of-pocket spending associated with travel for medical treatment, home modifications, and caretaking expenses, including loss of employment income, was not included but can add considerably to the out-ofpocket burden. Third, the inclusion of premium costs paid by insured people might have made comparisons of medical spending between the insured and the uninsured more balanced. Finally, the Medicare Part D prescription drug benefit did not begin until6,sowewereunabletoassesstheimpactofthisnewbenefitandthecoverage gap (the doughnut hole) on seniors out-of-pocket costs. Given the rising prevalence in chronic conditions among adults over age forty-four and the fact that many (but not all) chronic conditions can be attributed to poor lifestyle habits or better controlled with improved lifestyles, health insurance benefit redesign and health care reform should include incentives for people to adopt lifestyle practices that reduce chronic condition risk and improve health. Employers are increasingly recognizing the value of wellness programs and making them available to employees. However, more dramatic and systematic efforts are needed to induce a societal shift where primary and secondary prevention is considered a basic benefit and healthy lifestyles are the cultural norm. The authors thank Lauren McGivern for her assistance in preparing this manuscript. NOTES. V. Goff, Consumer Cost Sharing in Private Health Insurance: On the Threshold of Change, Report no. 798 (Washington: National Health Policy Forum, ). January/February 9

11 . Ibid. 3. J.C. Robinson, Renewed Emphasis on Consumer Cost Sharing in Health Insurance Benefit Design, Health Affairs (): w39 w5 (published online March ;.377/hlthaff.w.39).. M.E. Chernew and J.P. Newhouse, What Does the RAND Health Insurance Experiment Tell Us about the Impact of Patient Cost Sharing on Health Outcomes? American Journal of Managed Care, no. 7 (8):. 5. See, for example, C.Y. Lu et al., Interventions Designed to Improve the Quality and Efficiency of Medication Use in Managed Care: A Critical Review of the Literature 7, BMC Health Services Research 8 (8):75;T.B.Gibson,R.J.Ozminkowski,andR.Z.Goetzel, TheEffectsofPrescriptionDrugCostSharing: A Review of the Evidence, American Journal of Managed Care, no. (5): 73 7; S.B. Soumerai et al., Effects of a Limit on Medicaid Drug-Reimbursement Benefits on the Use of Psychotropic Agents and Acute Mental Health Services by Patients with Schizophrenia, New England Journal of Medicine 33, no. (99): ; P.B. Landsman et al., Impact of Three-Tier Pharmacy Benefit Design and Increased Consumer Cost-Sharing on Drug Utilization, American Journal of Managed Care, no. (5): 6 68; J. Hsu et al., Unintended Consequences of Caps on Medicare Drug Benefits, New England Journal of Medicine 35, no. (6): ; anda. Chandra, J. Gruber, and R. McKnight, Patient Cost-Sharing, Hospitalization Offsets, and the Design of Optimal Health Insurance for the Elderly, NBER Working Paper no. W97 (Cambridge Mass.: National Bureau of Economic Research, March 7). 6. W. Hwang et al., Out-of-Pocket Medical for Care of Chronic Conditions, Health Affairs, no. 6 (): J. Cohen, Design and Methods of the Medical Expenditure Panel Survey Household Component, MEPS Methodology Report no., Pub. no. 97-6, 997, publications/mr/mr.pdf (accessed 3 October 8). 8. Hwang et al., Out-of-Pocket Medical, See Agency for Healthcare Research and Quality, 996 Medical Conditions, MEPS Data Documentation HC- 6, Pub. no. 99-DP6 (Rockville, Md.: AHRQ, 999).. Hwang et al., Out-of-Pocket Medical ; and Bureau of Labor Statistics, Consumer Price Index Tables, (accessed October 8).. Decision rules for insurance status were developed based on four age categories: over age 65, under age 65, exactly 65, and newborn. A detailed description of the decision-making hierarchy to assign insurance status can be obtained by contacting Wenke Hwang at whwang@wfubmc.edu.. Poverty status was classified in MEPS as follows: poor family income below or equal to the federal poverty level; near-poor 5 percent of poverty; low income 6 percent of poverty; middle income percent of poverty; and high income more than percent of poverty. 3. Hwang et al., Out-of-Pocket Medical.. Source of National Health Expenditures data to calculate percentage increase in NHE from 995 to 6 was Centers for Medicare and Medicaid Services, National Health Expenditures by Type of Service and Source of Funds: Calendar Years 6 96, available at CMS, National Health Expenditure, Data, Historical, (accessed 3 October 8). 5. Hwang et al., Out-of-Pocket Medical. 6. I.B. Ahluwalia and J. Bolen, Lack of Health Insurance Coverage among Working-Age Adults, Evidence from the Behavioral Risk Factor Surveillance System, 993 6, Journal of Community Health 33, no. 5 (8): A. Davidoff and G.M. Kenney, Uninsured Americans with Chronic Conditions: Key Findings from the National Health Interview Survey (Princeton, N.J.: Robert Wood Johnson Foundation, 5). 8. Chernew and Newhouse, What Does the RAND Health Insurance Experiment Tell Us? 9. K. Baicker and A. Chandra, Myths and Misconceptions about U.S. Health Insurance, Health Affairs 7, no. 6 (8): w533 w53 (published online October 8;.377/hlthaff.7.6.w533).. G. Claxton et al., Health Benefits in 8: Premiums Moderately Higher, while Enrollment in Consumer- Directed Plans Rises in Small Firms, Health Affairs 7, no. 6 (8): w9 w5 (published online September 8;.377/hlthaff.7.6.w9). HEALTH AFFAIRS ~ Volume 8, Number 5