1 - This and subsequent MEPS-based tables on the non-Medicare population are limited to nonelderly people without Medicare. They omit the very small number of elderly people—an estimated 150,000—who are not covered by Medicare.
2 - Appendix B contains a further discussion of use rates and how users are treated in this report.
3 - The difference in the price of prescriptions purchased by covered people compared to those purchased by noncovered people is only statistically significant at the five percent level on a one-tail test.
4 - In some cases, people with supplemental insurance may even be paying premiums for the addition of a drug benefit that cost more on average than the maximum value of the drug benefit.
5 - Note that not every beneficiary with drug coverage obtained that coverage through the same source that provided his or her primary Medicare supplemental coverage.
6 - However, Medicaid beneficiaries’ spending, utilization, and out-of-pocket spending is not significantly different from the next closest group in each case: their spending is only different from spending by those with employer-sponsored coverage on a one-tail test, their utilization is not significantly different from beneficiaries with other public insurance, and their out-of-pocket spending is not significantly different from beneficiaries enrolled in an HMO.
7 - Differences among beneficiaries enrolled in Medicare HMOs are not statistically significant, primarily because of the small number of beneficiaries in HMOs without drug coverage.
8 - Income is given as a percentage of the federal poverty threshold, which differs by age and household size. For beneficiaries over 65, these calculations use a poverty threshold of $7,525 for an individual, and $9,491 for those who lived with others. For beneficiaries under 65 they use $8,163 for an individual, and $10,564 for those who lived with others.
9 - In the income range between 136 percent and 150 percent of poverty, the difference in spending between beneficiaries with and without coverage is not statistically significant.
10 - However, the smallest difference, between users with and without coverage with incomes between 100 and 200 percent of poverty, is statistically significant only on a one-tail test.
11 - For persons without coverage, the difference in spending between those with good health status and those with very good health status is not statistically significant, nor is the difference between persons with fair health and those with poor health.
12 - Among beneficiaries with the same health status, the spending difference between those with and without coverage is statistically significant for all beneficiaries except those in excellent health.
13 - The difference between spending by covered beneficiaries aged 65 to 69 and those aged 85 and over is not statistically significant.
14 - Hellinger (1995), PPRC (1996), Riley (1996), Hamilton (1999), and Call et al (1999); see bibliography in Appendix A for detailed citations.
15 - Ettner (1997), Wolfe and Goddeeris (1991), and Atherly (1999) found evidence of selection. Cartwright, Hu, and Huang (1992), Hurd and McGarry (1997), and Lillard et al. (1999) did not. See bibliography in Appendix A for detailed citations.
16 - Atherly, A. (1999). “The Effect of Medicare Supplemental Insurance on Medicare Expenditures.” Department of Health Policy and Management. School of Public Health. Emory University. Atlanta, GA. Unpublished Paper.
17 - None of the measures used are direct indicators of the need for prescription drugs. However, that drug spending rises with deteriorating health for both covered and noncovered people suggests that the measures are at least strongly correlated with prescription drug need.
18 - The experimental design allowed researchers to eliminate the effects of self-selection by randomly assigning participants to plans that required various levels of copayment, ranging from full first-dollar coverage to 95 percent coinsurance (to simulate non-coverage).
19 - Brook, Robert, et al., “Does Free Care Improve Adults’ Health? Results from a Randomized Controlled Trial,” NEJM, 309 (23), Dec. 8, 1983, 1426-1434; Lohr, Kathleen, et al., “Use of Medical Care in the RAND Health Insurance Experiment: Diagnosis and Service-Specific Analyses in a Randomized Controlled Trial,” Medical Care 24(9) (Supplement), Sept. 1986, S74-S77; Manning, Willard G., et al. "Health Insurance and the Demand for Medical Care: Evidence from a Randomized Experiment," American Economic Review, 77(3), June 1987, 251-277.
20 - Stuart and Grana (1998), Stuart and Zacker (1999), Ya-chen (1999), and Lillard, Rogowski, and Kingston (1999); see bibliography in Appendix A for detailed citations.
21 - Coulson NE. and Stuart, BC. (1995) “Insurance Choice and the Demand for Prescription Drugs.” Southern Economic Journal. 61(4): 1146-1157.
22 - General Accounting Office. Prescription Drugs and the Elderly: Many Still Receive Potentially Harmful Drugs Despite Recent Improvements, March 1996.
23 - The difference between covered and noncovered users in the next to highest quintile is significant only on a one tail test.
24 - A consumer unit comprises either: (1) all members of a household who are related by blood, marriage, adoption, or other legal arrangements; (2) a person living alone, or a person sharing a household who is financially independent; or (3) two or more persons living together who pool their income to make joint expenditure decisions. The reference person is the person named when the respondent is asked to "Start with the name of the person or one of the persons who owns or rents the home." It is with respect to this person that the relationship of the other consumer unit members is determined.
25 - Total reported spending for units with reported income below $15,000 actually exceeds $15,000. Data in the tables are for complete income reporters. These are units that provide values for at least one of the major sources of their income. However, even complete income reporters do not necessarily provide full accounting of their income. In addition, some units may spend more than their income by borrowing or by using up assets; the latter may be more likely for the elderly.
26 - Note that these figures cannot be compared to the MCBS and MEPS out-of-pocket numbers for individuals because the year is different, and many units have more than one individual.
27 - The MCBS Access to Care component for 1996 asked whether the respondent had failed to fill one or more prescriptions during the year, and why. By contrast, the NHIS question asks whether a respondent failed to receive a needed drug. This study does not include an analysis of the MCBS question because it does not directly capture people who might never have had a drug prescribed because they did not go to the doctor.
28 - The exception is for people with Medicaid, for whom the difference between income groups is not statistically significant.
29 - Note that the analysis included beneficiary use of drugs associated with hypertension, but some of these medications may have been prescribed for purposes other than treatment of hypertension. For example, diuretics could be prescribed for a number of purposes.
30 - However, Chapter 3 will show that people with and without coverage purchase a similar mix of brand-name and generic drugs. More analysis is needed on this issue.
31 - Copayments are not necessarily less burdensome than coinsurance. On a $35 prescription, 20 percent coinsurance would be $7, less than the copayments imposed by many plans. However, the elimination of deductibles probably reduced consumer liabilities considerably. In addition, coinsurance rises automatically with drug inflation, while insurers have to change copayments periodically to keep pace with inflation.
32 - HCFA(now known as CMS), based on the CPI and the GDP deflator.
33 - PhRMA, Drug Utilization And Managed Care, Sept. 1998.
"intro.pdf" (pdf, 23.11Kb)
"C1.pdf" (pdf, 75.87Kb)
"c2.pdf" (pdf, 169.02Kb)
"c3.pdf" (pdf, 92Kb)
"future.pdf" (pdf, 12.41Kb)
"appena.PDF" (pdf, 149.34Kb)
"appenb.pdf" (pdf, 27Kb)