QUICK SEARCH:   [advanced]
Author:
Keyword(s):
Year:  Vol:  Page: 

   

 

This Article
* Abstract Freely available
* Figures Only
* Reprint (PDF)
* Submit a response to this article
* Alert me when this article is cited
* Alert me when Comments are posted
* Alert me if a correction is posted
Services
* E-mail this article to a friend
* Similar articles in this journal
* Similar articles in PubMed
* Alert me to new issues of the journal
* Add to My Personal Archive
* Download to Citation Manager
*Reprints & Permissions
Citing Articles
* Citing Articles via HighWire
* Citing Articles via Google Scholar
Google Scholar
* Articles by Stuart, B.
* Articles by Briesacher, B.
* Search for Related Content
PubMed
* PubMed Citation
* Articles by Stuart, B.
* Articles by Briesacher, B.
Related Collections
* Managed Care - Medicare
* Medicare
* Pharmaceuticals

Drug Cost Growth

Dynamics In Drug Coverage Of Medicare Beneficiaries: Finders, Losers, Switchers

Bruce Stuart, Dennis Shea and Becky Briesacher

   Abstract
 
Although the vast majority of elderly Americans have the stability of basic Medicare benefits, Medicare alone offers no protection from the vicissitudes of the market for outpatient prescription drugs. This paper analyzes the sources and stability of prescription coverage maintained by Medicare beneficiaries in 1995 and 1996. The results show that fewer than half of all beneficiaries had continuous drug coverage over this period, while nearly a third gained, lost, or had spells without coverage.


As the debate over a Medicare prescription drug benefit intensifies, policymakers need sound information about current levels of drug coverage maintained by program beneficiaries. Although nearly 70 percent of all beneficiaries are reported to have drug coverage, the stability of that coverage has been questioned.1 Stability of health insurance has long been recognized as an important issue facing working Americans.2 Workers who have experienced spells without insurance are more likely to encounter access and cost problems, are more likely to postpone needed care, and are less satisfied with the health care that they do receive.

In a previous study we showed that one in five Medicare beneficiaries with drug coverage in 1996 had it for only part of the year and that these individuals had levels of drug use and spending that were well below those of persons with full-year drug coverage.3 In this paper we explore the sources of instability by examining the characteristics of persons who lost, found, or switched drug benefits between January 1995 and December 1996. The selection of a two-year time frame helps to distinguish persons who are true losers and gainers from those who move into and out of coverage over time. It also enables us to capture changes in drug coverage that occur at the beginning of the calendar year, when many organizations make benefit decisions. All previous studies of Medicare beneficiaries’ drug coverage have analyzed annual data and are thus blind to shifts in coverage that cross calendar years.

For beneficiaries whose drug coverage changed over the study time frame, we identify the sources of their drug benefits and analyze factors that might explain the shifts in coverage. Although the data do not permit us to distinguish voluntary behavior from forced decisions, we pay particular attention to factors such as personal income, geographic location, and burden of chronic illness, which may indicate problems in access to drug benefits. The results demonstrate that policymakers’ concerns about the stability of drug coverage for Medicare beneficiaries are well founded.

   Data Source And Methods
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
As in our prior work, the data source for this study is the Medicare Current Beneficiary Survey (MCBS).4 We used the latest available data to track the drug coverage patterns of a cohort of community-dwelling MCBS respondents represented in the survey for all of 1995 and 1996. This continuous sample is not strictly representative of the entire noninstitutionalized Medicare population because we excluded new Medicare enrollees and persons who died.5 However, statistical tests indicate that the characteristics of our cohort of 6,185 MCBS respondents closely match those of the general Medicare population as of 1996.6

To highlight the full spectrum of drug coverage dynamics, we classified cohort members into seven mutually exclusive categories that describe the duration and consistency of their prescription coverage: (1) continuous plans, (2) sequential plans, (3) lost coverage, (4) found coverage, (5) intermittent coverage, (6) some coverage but not reported, and (7) no coverage.7 Assignment into the categories was made according to responses to MCBS questions about Medicare supplemental insurance. Respondents were asked to list all active supplementary insurance plans, the plan’s beginning and end dates, and whether prescription drugs were covered.8 If the beneficiary indicated drug coverage, we assumed that it was maintained throughout the entire duration of the source plan or until the beneficiary reported that the coverage had ended. Respondents were also asked who paid for each filled prescription. A small number of respondents reported third-party drug payments but no drug coverage.

To simplify the analysis, we assumed that prescription coverage extended for the full month in any month when a beneficiary lost, found, or switched coverage. These assumptions impart a slight upward bias to the drug coverage rates reported here. Finally, we used regression analysis to determine the independent contribution of personal characteristics to an individual’s classification into a drug coverage group.9 Only associations that are significant at conventional levels of statistical significance (p < .05) are reported here.

   Characteristics Of Beneficiaries By Drug Coverage Status
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
Exhibit 1Go shows the proportion of beneficiaries falling into the seven prescription coverage categories grouped together to denote cases of continuous coverage, gaps in coverage, or no coverage. Coverage rates for cohort members in calendar year 1995 are slightly lower than the coverage levels reported in our earlier analysis for CY 1996, which is consistent with published trends.10 The exhibit also shows coverage rates measured over the entire two-year period. The reason for taking the longer time perspective is readily apparent from these statistics.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
EXHIBIT 1 Prescription Drug Coverage Rates For Noninstitutionalized Medicare Beneficiaries In 1995 And 1995–96, By Composition Of Coverage Classes

 
The percentage of beneficiaries maintaining continuous prescription benefits (in either the same plan or successive plans) over two years is 46.3, six percentage points lower than those doing so in a single year (52.7 percent). This population comprises two groups: a dominant class of beneficiaries with coverage from the same plan or plans, and a smaller group of persons with successive plans and no break in coverage.11 The number of beneficiaries having continuous coverage through multiple, part-year plans more than doubles when we follow them from one to two years.

Nearly one in three beneficiaries had gaps in their prescription coverage at some point during the two years, compared with 16.7 percent in 1995 (Exhibit 1Go). The cohort population with discontinuous coverage is spread among four distinct groups. The largest entered 1995 with no prescription coverage and ended 1996 with it. We define these as prescription coverage "finders." Actually, 23.4 percent of the entire cohort "found" coverage during this two-year period if one counts beneficiaries with intermittent and successive coverage. Finders outnumbered losers no matter how one counts them. Beneficiaries with no coverage at all dropped seven percentage points between the one-year and two-year views.

Exhibit 2Go depicts the demographic characteristics of beneficiaries with continuous prescription coverage, gaps, and no coverage. The age structure of the coverage groups was roughly similar, although a much higher proportion of the "no coverage" group were age eighty or older than was true for the continuously covered group. There are other evident differences, many of which are significant in bivariate statistical tests. However, as is often the case, demographic and health factors tend to cluster together (for example, low income, advanced age, and number of chronic health conditions). The regression models comparing the three groups included all of the variables shown in Exhibit 2Go except source of prescription coverage, which is undefined for persons with no coverage.


View this table:
[in this window]
[in a new window]
EXHIBIT 2 Characteristics Of Noninstitutionalized Medicare Beneficiaries, By Prescription Drug Coverage Status In 1995 And 1996

 
Strongest predictors. The strongest predictors of prescription drug coverage are geographic residence, income, and the burden of chronic illness. Beneficiaries living in the western United States had the highest coverage rates for both continuous and partial-year benefits, and beneficiaries in the Midwest, the lowest. Independent of all other factors, where one lives can boost the probability of having drug coverage by up to 300 percent. Residence in a state with a pharmaceutical assistance program increases the probability of continuous coverage by nearly 60 percent and that of partial coverage by almost 20 percent. Residing in an urban area increases the likelihood of having either continuous or part-year coverage by almost 90 percent compared with rural residence. As one would expect, high income is a positive predictor of having drug coverage, and low income, of being without it. There is also a strong positive association between drug coverage and the burden of chronic illness. The probability of having continuous drug coverage increases by approximately 20 percent for each additional chronic condition reported. Other factors such as age, race, and ethnicity are insignificant in the multivariate analyses.

Importance of coverage source. Regression results comparing persons with continuous coverage and coverage gaps produced some provocative and unexpected results (since every individual in both groups had some source of drug benefits, we included variables for insurance type in this model). Although about half of all beneficiaries with continuous coverage obtained it from an employer-sponsored plan (see Exhibit 2Go), persons with drug coverage under such plans had a higher risk of suffering a gap in coverage than did those with any other source of prescription benefits, except for individual Medigap plans. In fact, the source of drug benefits is more important in explaining which beneficiaries experienced gaps in coverage than all of the other variables listed in Exhibit 2Go combined.12 As might be expected, persons with public sources of prescription benefits including Medicaid, Qualified Medicare Beneficiary Plus and Specified Low-Income Medicare Beneficiary Plus programs (QMB Plus/SLMB Plus), and other public plans exhibited the most stable prescription benefits.13 In the multivariate model, beneficiaries with individual Medigap policies were almost twice as likely to have gaps in prescription drug coverage than were persons with employer-sponsored drug benefits, all else being equal.

Finders, losers, and switchers. We also performed regression analysis to ascertain differences in characteristics of beneficiaries who found, lost, and had intermittent drug coverage. Finders tended to be younger, live in the West and in urban areas, and have lower incomes compared with those who have no coverage. There is no evidence that health status or burden of chronic illness played a major role in determining who obtained prescription drug coverage in 1995 and 1996.

Comparing losers with those having continuous plans paints a very different picture. The 7.4 percent of beneficiaries who lost drug coverage over the biennium were older, less likely to have poverty-level incomes, more likely to reside in the Midwest, and have a much lower burden of chronic disease. The type of prescription drug coverage held before the loss is an important factor. Losers were much more likely to have Medigap or Medicare HMO coverage than other sources of coverage. Persons with public coverage were much less likely to lose drug benefits.

The 8 percent of the cohort with intermittent drug coverage over the biennium were much more likely to have had a spell of coverage from a private plan than from a public program. This is consistent with the finding noted above that enrollment in public programs is predictive of continuity in coverage. Regression analysis showed that males were about 25 percent more likely than females were to have intermittent drug coverage. Persons with annual incomes below $10,000 were 2.5 times as likely to have intermittent coverage as persons with incomes exceeding $50,000 were. Persons residing in a state with a pharmaceutical assistance program were about 40 percent less likely than residents of other states were to have intermittent drug coverage.

Nonreporters. The final category of beneficiaries with part-year coverage—those with evidence of some coverage but not reporting it—represent an anomalous group. Compared with the continuously covered, the 4 percent of beneficiaries in this category reported being in poorer health and having many more concurrent chronic conditions (35 percent reported four or more chronic conditions compared with just 23 percent for the continuously covered). Other significant factors predictive of this class are sex (males were more likely to not report coverage), residence in rural areas, and annual incomes below $30,000.

   Changes In Drug Coverage Over The Biennium
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
Exhibit 3Go presents a two-period analysis of drug coverage status in 1995 and 1996.We see that 84 percent of the beneficiaries who had continuous prescription coverage plans in 1995 maintained them through 1996. In other words, stability in the past strongly predicts stability in the future. The 16 percent who failed to maintain continuous plans were sprinkled among the other coverage groups. Nearly 8 percent lost coverage in 1996 (1.8 percent lost it in January and thus are counted as not having coverage at all during 1996; the 5.9 percent reported here lost it at some point after January 1996). Almost 5 percent were able to maintain coverage throughout 1996 with sequential plans, and a smattering had intermittent coverage in 1996 or failed to report coverage that they had (based on claims evidence).


View this table:
[in this window]
[in a new window]
EXHIBIT 3 Changes In Prescription Drug Coverage Status From 1995 To 1996 For Noninstitutionalized Medicare Beneficiaries

 
Persons in the other coverage status groups were considerably more dynamic in their behavior than was the continuously covered group (Exhibit 3Go). About 70 percent of the sequentially covered group in 1995 went on to change coverage again in 1996. Only 8 percent of this group found a plan that provided continuous coverage throughout 1996. By contrast, very few beneficiaries who had intermittent coverage in 1995 did so again in 1996. Most either obtained continuous coverage in 1996 or lost it for the whole year. About half of the beneficiaries who failed to report coverage in 1995 did so again in 1996. About 28 percent reported finding coverage from continuous, successive, or part-year plans, and 20 percent had no coverage in 1996. The final group—those with no coverage whatsoever in 1995—were nearly as stable as the continuously covered.

Exhibit 4Go examines changes in drug coverage by tracking movements into and out of the various sources of drug benefits available to Medicare beneficiaries. Here we see that 29.1 percent of the study cohort had employer-sponsored drug coverage at the beginning of the biennium and 28.9 percent had it at the end. Focusing only on the difference might lead one to conclude that employer coverage is highly stable. In fact, that is not the case at all. One-tenth of the sample reported adding an employer drug plan, and an additional 1.3 percent switched into such a plan. More than offsetting these changes were the 12.9 percent of beneficiaries who either dropped an employer-sponsored drug benefit plan outright or switched to another kind of coverage.14


View this table:
[in this window]
[in a new window]
EXHIBIT 4 Medicare Beneficiaries Adding, Dropping, And Switching Prescription Drug Benefits, By Source Of Coverage, 1995–1996

 
Other private sources of drug coverage are also unstable. The slight increase in proportion of beneficiaries reporting Medigap drug benefits from January 1995 to December 1996 belies the turnover evident in between. Overall, only 48 percent of beneficiaries who began with an individual Medigap drug benefit ended with one. The retention rate for health maintenance organization (HMO) enrollees fell between that for beneficiaries with employer coverage and Medigap policies. Three-quarters of the January 1995 HMO enrollees also reported an HMO drug benefit in December 1996, and 89 percent of these maintained one continuously but not necessarily from the same plan.

Public programs exhibited much lower turnover rates than private sources of drug coverage did. Medicaid enrollment rose slightly over the biennium, and just 16 percent of January recipients failed to maintain Medicaid coverage through to December of the next year. Beneficiaries with drug coverage through a QMB Plus or SLMB Plus program were less likely to maintain continuous coverage from this source (79 percent did so). New recipients account for most of the change in enrollments in other public plans offering drug coverage.

   Tracking Drug Coverage Through The Biennium
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
We wrap up our analysis by examining month-to-month changes in drug coverage (Exhibit 5Go). Here the upward trend in drug coverage rates over the biennium can be seen in relation to much bigger—and largely compensating—shifts in and out of plans. Over the entire biennium the monthly rate at which beneficiaries added prescription drug plans exceeded that of plan drops (hence the increase in drug coverage rates over time). But in between, plan adds and drops followed a phased pattern with distinct spikes in the early spring and summer months. The biggest spike of all occurred between December 1995 and January 1996. All told, 9.6 percent of the entire study cohort experienced a change in drug coverage in that month.


Figure 2
View larger version (16K):
[in this window]
[in a new window]
EXHIBIT 5 Percentage Of Noninstitutionalized Medicare Beneficiaries Adding, Dropping, Or Switching Prescription Coverage, By Month, 1995–1996

 
We analyzed the composition of these month-to-month changes, to identify the sources and targets of prescription drug coverage. The bubble in coverage rates during the early summer months of 1995 was driven by Medicare beneficiaries adding drug benefits from employer-sponsored and Medigap plans in May and June. Coverage losses from employer plans and Medigap policies were also concentrated in the summer months but were less frequent and occurred later in the season. No seasonal patterns were evident in coverage switches involving Medicaid, QMB Plus/SLMB Plus, or other public plan types in 1995. The spike in plan movements between December and January was dominated by changes in employer and HMO drug coverage: 37 percent of all persons adding drug benefits that month obtained them through an employer plan, compared with 19 percent from a Medicare HMO. Of persons dropping drug coverage in January, 23 percent left an employer plan, compared with 47 percent leaving HMO coverage. The most common coverage switch in January involved dropping an employer plan and adding HMO coverage.15 In 1996 the picture began to change. Fewer beneficiaries added drug benefits from employer plans and HMOs than in 1995, and more purchased drug coverage from Medigap plans. Almost half of all those losing drug benefits in 1996 dropped them (or were dropped) from an employer plan.

   Study Limitations
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
All of the findings presented in this paper are derived from a survey sample and are thus subject to the variability inherent in any such sample. We have employed conservative levels of statistical significance to interpret our findings, but there is always a small chance that a particular finding labeled as significant is really an artifact. Strictly speaking, the study results cannot be extrapolated to the entire Medicare population because we systematically excluded new Medicare entrants, persons who died, and nursing home residents from our cohort. Although we have no reason to believe that such persons were immune to the dynamic forces that shaped prescription drug coverage patterns for the rest of the Medicare population, we have no data to support that supposition.

One major limitation of the MCBS is the lack of data on scope of benefits for those with Medicare supplemental policies. We can be reasonably assured that MCBS respondents who had stable Medicaid coverage had generous drug benefits throughout the study period. Employer plans are also thought to be relatively generous, at least compared with the three standard Medigap policies that provide drug coverage. Medigap plans H, I, and J pay only 50 percent of drug costs above a $250 annual deductible and cap total benefits at either $1,250 or $3,000. Almost all Medicare HMO plans offered drug coverage in 1995 and 1996, but the benefits varied widely. This variability in depth of coverage adds further heterogeneity to an already unstable pattern in coverage rates.

A final shortcoming of the MCBS is its lack of timeliness. The lag of several years puts policy analysts in the unenviable position of basing projections on historical trends known not to reflect the present and unlikely to persist in the future.

   Discussion And Policy Implications
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 
Our findings paint a very different picture of Medicare beneficiaries’ drug coverage than that described by other researchers. We do find, as others have, that coverage rates improved between 1995 and 1996.16 Underneath the surface, however, is a highly dynamic if not chaotic market for drug benefits. When viewed over this two-year period, the number of beneficiaries finding, losing, and switching drug coverage is almost as large as the number with stable coverage.17

What does this mean for policymakers? The answer hangs on two major questions. First, is the instability documented here real? Second, is the instability voluntary or forced?

On the first question, the technical answer is that if you follow beneficiaries long enough, most are likely to have had some period with prescription drug coverage. The fact that the percentage of beneficiaries with no coverage drops from 30 percent to 23 percent when the window of observation extends from twelve to twenty-four months strongly points in that direction. However, the percentage of beneficiaries finding and losing coverage also increases with the twenty-four-month window. Were it possible to look beyond twenty-four months, this pattern might change. But even if losers, for instance, eventually become finders, their gaps in drug coverage are real enough to engender public concern.

The issue of forced versus voluntary changes in drug coverage is really about affordability and access to insurance. We find that beneficiary characteristics—at least those captured in the MCBS—play a relatively small role in explaining plan-switching behavior. The single most important factor is the source of coverage itself. In all of our regression analyses, public plans offer more stable prescription benefits than do private plans. Not surprisingly, given the limited benefits packages, beneficiaries with Medigap policies are the least likely to have stable drug coverage. However, contrary to expectations, employer-sponsored plans were not a reliable source of stable drug benefits either. This finding is particularly important given the erosion of retiree coverage in the years since the data for this study were compiled.18 Drug benefits available through Medicare HMOs have also declined in recent years.19

Next to the source of prescription benefits, where beneficiaries live plays a major role in the stability of coverage. We found that residents of the Midwest are much less likely to have stable prescription coverage than beneficiaries in any other region of the country and are at greater risk of losing the coverage they do have. Residents in the West have the most stable drug benefits. Residing in a state with a pharmaceutical assistance program greatly boosts the probability of having stable prescription drug coverage. Likewise, living in urban areas raises the likelihood of finding and keeping drug benefits. These associations clearly reflect underlying differences in access to good prescription drug coverage.20 The accident of geography may ultimately provide the strongest argument for enacting a uniform Medicare drug benefit.

Instability in currently available prescription plans adds fuel to the argument that Medicare should include drug coverage. Stable drug coverage would help to alleviate the economic risk that beneficiaries now face when confronted with changes in prescription drug coverage and premium rates in their Medicare supplemental plans. Even if most of the coverage switches we have observed are "voluntary" in the sense that beneficiaries have a choice to retain, add, or drop coverage, the decision to maintain stable prescription drug coverage may not be a realistic option for many of them.

   Editor's Notes
 
Bruce Stuart is director, Peter Lamy Center on Drug Therapy and Aging, and Parke-Davis Professor of Geriatric Pharmacotherapy, at the University of Maryland School of Pharmacy. Dennis Shea is a professor in the Pennsylvania State University Department of Health Policy and Administration. Becky Briesacher is a doctoral candidate at the University of Maryland School of Pharmacy.

The authors thank Steve Long, Marilyn Moon, Susan Raetzman, and Judy Wagner for helpful comments on an earlier draft of this paper. Cheryl Fahlman and Steve Maczuga provided valuable programming assistance. Research support for this work was provided by the Commonwealth Fund. The views presented here are those of the authors and not necessarily those of the Commonwealth Fund, its directors, officers, or staff.

   NOTES
 Top
 Data Source And Methods
 Characteristics Of Beneficiaries...
 Changes In Drug Coverage...
 Tracking Drug Coverage Through...
 Study Limitations
 Discussion And Policy...
 NOTES
 

  1. J.A. Poisal and G.J. Chulis, "Medicare Beneficiaries and Drug Coverage," Health Affairs (Mar/Apr 2000): 248–256; and U.S. Department of Health and Human Services, Report to the President: Prescription Drug Coverage, Spending, Utilization, and Prices (Washington: DHHS, April 2000).
  2. R. Bennefield, Who Loses Coverage and for How Long? Census Bureau Pub. no. P70-54 (Washington: U.S. Department of Commerce, 1996); C. Copeland, Characteristics of the Non-Elderly with Selected Sources of Health Insurance and Lengths of Uninsured Spells, EBRI Issue Brief no. 198 (Washington: Employee Benefit Research Institute, 1998); P. Farley, "Who Are the Underinsured?" Milbank Memorial Fund Quarterly: Health and Society 63, no. 3 (1985): 476–503; C. Schoen et al., Working Families at Risk: Coverage, Access, Cost, and Worries (New York: Commonwealth Fund, 1998); P. Short, "New Estimates of the Underinsured Younger than Sixty-five Years," Journal of the American Medical Association 274, no. 16 (1995): 1302–1306[Abstract/Free Full Text]; and C. Schoen and C. DesRoches, "Uninsured and Unstably Insured: The Importance of Continuous Coverage," Health Services Research 35, no. 1 (2000): 187–206.[Medline]
  3. B. Stuart, D. Shea, and B. Briesacher, Prescription Drug Costs for Medicare Beneficiaries: Coverage and Health Status Matter, Issue Brief (New York: Commonwealth Fund, January 2000).
  4. Extensive information about the MCBS is available at <www.hcfa.gov/mcbs>.
  5. The study inclusion criteria were as follows: (1) MCBS respondents with Medicare enrollment dates after 31 December 1994 and continuously covered through 31 December 1996; (2) eligible for all six survey rounds of 1995 and 1996; and (3) included in the community-dwelling sample in first round of 1995. Respondents were excluded if they (1) had gaps in Medicare entitlement, (2) were included in the MCBS institutional sample in 1995, (3) died in 1995 or 1996, or (4) voluntarily dropped out of the MCBS survey during the study period. We considered tracking the same individuals for more than two years, but the sample size decreased too much to permit meaningful analysis.
  6. There were no statistically significant differences between the cohort and the full sample of noninstitutionalized MCBS respondents in the following variables: sex, race, marital status, residence, income, and prescription drug coverage. The cohort members were slightly older and less likely to report fair/poor health status than was true for the general Medicare population in 1996.
  7. We defined our drug coverage categories as follows: (1) continuous plans: persons had prescription drugs from one or more plans, any of which provides continuous coverage over the reporting period; (2) sequential plans: persons were continuously covered by successive or overlapping plans; (3) lost coverage: persons began the reporting period with one or more plans that cover drugs but did not end the period with coverage; (4) found coverage: persons began the reporting period without drug coverage but ended the period with one or more plans that cover drugs; (5) intermittent coverage: persons with gaps in coverage both lost and found drug coverage during the reporting period; (6) some coverage but not reported: persons did not report having any plan with prescription drug coverage, but their payment records show some third-party payments; and (7) no coverage: persons did not report having a plan with prescription drug coverage, and their payment records do not show any third-party drug payments.
  8. For public programs we assigned drug coverage to all beneficiaries entitled to full Medicaid benefits and to recipients of "other public programs" including Department of Veterans Affairs (VA) and state pharmaceutical assistance programs. Persons enrolled in the Qualified Medicare Beneficiary (QMB) and Specified Low-Income Medicare Beneficiary (SLMB) programs were assigned drug coverage if they had any evidence of a Medicaid-paid prescription claim; otherwise, they were classified as not having drug coverage.
  9. The models were estimated with multinomial logistic regression correcting for the complex sample design employed by the MCBS. Some models were restricted to beneficiaries with evidence of drug benefits to examine the source of coverage variables.
  10. See Poisal and Chulis, "Medicare Beneficiaries and Drug Coverage"; and Stuart et al., Prescription Drug Costs.
  11. About 1 percent of the larger group of beneficiaries had more than one continuous source of drug coverage.
  12. The model without variables representing source of drug coverage explained only 1.5 percent of the variance between those with gaps and continuous coverage. Adding variables for source of coverage increased the explained variance to 6.5 percent.
  13. The "Plus" signifies states that provide full Medicaid benefits (including drug overage) to QMB/SLMB recipients.
  14. A small number of the changes recorded during the biennium were by persons with more than one employer plan, so the net effect of adds, drops, and switches does not precisely reconcile the coverage rates in January 1995 with December 1996.
  15. This finding may well be an artifact of the way MCBS classifies enrollees in Medicare HMOs. Respondents are classified as Medicare HMO enrollees even if the employer paid their HMO premiums.
  16. See Poisal and Chulis, "Medicare Beneficiaries and Drug Coverage."
  17. There are actually more beneficiaries who changed drug coverage than did not when one considers that 6–8 percent of beneficiaries with continuous prescription drug coverage also added and/or dropped other sources of drug coverage in 1995 and 1996.
  18. See DHHS, Report to the President, chap. 1, for a good discussion of recent literature on retiree health coverage.
  19. Ibid.
  20. According to the Medicare Payment Advisory Commission, just 64 percent of Medicare beneficiaries now reside in areas served by Medicare+Choice plans. Medicare Payment Advisory Commission, Report to Congress: Medicare Payment Policy (Washington: MedPAC, March 2000), chap. 5.


Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati    What's this?


This article has been cited by other articles:


Home page
JAMAHome page
J. M. Madden, A. J. Graves, F. Zhang, A. S. Adams, B. A. Briesacher, D. Ross-Degnan, J. H. Gurwitz, M. Pierre-Jacques, D. G. Safran, G. S. Adler, et al.
Cost-Related Medication Nonadherence and Spending on Basic Needs Following Implementation of Medicare Part D
JAMA, April 23, 2008; 299(16): 1922 - 1928.
[Abstract] [Full Text] [PDF]


Home page
Journal of Health Politics, Policy and LawHome page
M. Schlesinger and J. S. Hacker
Secret Weapon: The "New" Medicare as a Route to Health Security
Journal of Health Politics Policy and Law, April 1, 2007; 32(2): 247 - 291.
[Abstract] [PDF]


Home page
J. Am. Soc. Nephrol.Home page
U. D. Patel and M. M. Davis
Falling into the Doughnut Hole: Drug Spending among Beneficiaries with End-Stage Renal Disease under Medicare Part D Plans
J. Am. Soc. Nephrol., September 1, 2006; 17(9): 2546 - 2553.
[Abstract] [Full Text] [PDF]


Home page
Med Decis MakingHome page
R. R. Cline and K. Gupta
Drug Benefit Decisions among Older Adults: A Policy-Capturing Analysis
Med Decis Making, May 1, 2006; 26(3): 273 - 281.
[Abstract] [PDF]


Home page
Arch Intern MedHome page
J. H. Gurwitz and P. Rochon
Improving the Quality of Medication Use in Elderly Patients: A Not-So-Simple Prescription
Arch Intern Med, August 12, 2002; 162(15): 1670 - 1672.
[Full Text] [PDF]


Home page
Am. J. Public HealthHome page
M. B. Artz, R. S. Hadsall, and S. W. Schondelmeyer
Impact of Generosity Level of Outpatient Prescription Drug Coverage on Prescription Drug Events and Expenditure Among Older Persons
Am J Public Health, August 1, 2002; 92(8): 1257 - 1263.
[Abstract] [Full Text] [PDF]


Home page
NEJMHome page
J. H. Lewis, M. Schonlau, J. A. Munoz, S. M. Asch, M. R. Rosen, H. Yang, and J. J. Escarce
Compliance among Pharmacies in California with a Prescription-Drug Discount Program for Medicare Beneficiaries
N. Engl. J. Med., March 14, 2002; 346(11): 830 - 835.
[Abstract] [Full Text] [PDF]



Home | Current Issue | Archives | Topic Collections | Search | Blog | Subscribe | Contact Us | Help

© 2001-2001 Project HOPE–The People-to-People Organization
Terms and Policies