The centerpiece of the Medicare Modernization Act of 2003 was a new drug benefit for enrollees, called Part D Medicare Part D began voluntary enrollment in 2006, with benefits available through supplemental stand-alone prescription drug plans or bundled as part of a Medicare Advantage supplemental plan At the end of the 2006 Part D enrollment period, 38.2 million Medicare beneficiaries had some source of drug coverage, including 22.5 million enrolled in Part D plans* Approximately 7% of seniors lacked drug coverage as of 2006, compared with 24% in 2004 Heiss et al – Retirement Perspectives Survey paper published in Aug. 2006 in Value in Health suggested that the healthy were enrolling in Part D at lower rates. Levy and Weir – Health and Retirement Study paper published Jan. 2009 in the NBER working papers series – 9,321 Medicare beneficiaries ≥ 65 years who completed core interviews in both 2004 and 2006 *U.S. Department of Health and Human Services, Press Release, June 14, 2006, http://hhs.gov/news/press/2006pres/20060614.html (accessed March 25, 2009)
About 6% of those covered by Medicare and in-scope in 2006 were missing information on Part D enrollment ( “refused, don’t know, or not ascertained”).
Income defined as family income relative to Federal poverty level (FPL); <200% of FPL = low income, 200-399% = middle income, 400% or more = high income. Premiums were excluded because all of our sample paid the same amount in Medicare monthly premiums, and private supplemental insurance was used as a variable in the regression.
Model variables: Age (65+ vs. under 65) Sex (male vs. female) Race/ethnicity (white vs. non-white) Region (South, Midwest, West vs. Northeast) Location (urban vs. rural) Marital status (married vs. unmarried) Education (12 years, 13+ years vs. <12 years) Income (high/middle vs. low) Physical SF-12 score (low or medium vs. high) Supplemental insurance coverage (Medigap only, Medigap + employer, employer only, other coverage vs. none) 2005 OOP drug expenditures categorized as <$1000, $1000-2500, $2500-5000, or >$5000
Out of 1,436 persons who met inclusion criteria, 657 (45.8%) enrolled in Part D during 2006 Compared to the no-Part D group, the Part D group: Was slightly older and more likely to be female Had more non-whites Had more unmarried individuals
Compared to the no-Part D group, the Part D group: Had more rural residents Was slightly less educated Was poorer
SF-12 scores range from 0-100, with higher scores indicating better health. Mean SF-12 physical component score for ages 70-79 is approximately 41.
In 2005, the group that enrolled in Part D in 2006 had: More beneficiaries with no private supplemental insurance Fewer beneficiaries with employer-based coverage only More beneficiaries with Medigap coverage only Almost 92% of the beneficiaries who had employer-based coverage in 2005 maintained that coverage in 2006.
Less than 12% of the overall sample had OOP drug expenditures ≥$2500 in 2005 The proportion was twice as high in the Part D group as in the no-Part D group (16.7% vs. 8.0%) Median OOP drug expenditures were $745 in the Part-D group and $409 in the no-Part-D group
In adjusted analyses, beneficiaries were less likely to enroll in Part D if they: Had Medigap plus employer or employer-only supplemental insurance Significant positive predictors of Part D enrollment were: Low income Having Medigap supplemental insurance only OOP drug expenditures between $1000-2500 or more than $5000 in 2005
Our results are in line with those of Heiss et al (2007), Neuman et al (2007), and Levy and Weir (2009) in terms of the proportion of elderly Medicare beneficiaries with Part D coverage in 2006. Others have suggested that employers might stop offering retiree benefits with the advent of Part D, but our study did not find this to be true, at least for 2006. This is in line with Levy and Weir’s findings from the HRS and Gabel et al’s survey of employers published in 2008 suggesting that there was very little crowd-out associated with the introduction of part D – people transitioning from employer coverage to Part D would likely have lost that coverage anyway. People with employer-sponsored supplementary health insurance (likely with prescription drug benefits) in 2005 also acted rationally by choosing not to enroll, whereas people with Medigap only in 2005 were not likely to have had drug benefits and rationally chose to enroll.
Creditable coverage is defined by Medicare as at least as good as the Part D Standard Plan. As of June 14, 2006, 69% of people age 65+ were enrolled in Part D, 24% had other creditable prescription drug coverage, and 7% had no creditable coverage (CMS data). Other unmeasured factors: distrust of the government or dissatisfaction with the Medicare program or healthcare system in general
Medicare Part D Podium 2009 05 13
Predictors of Enrollment in Medicare Part D: Are Beneficiaries Rational? Lisa M. Lines, MPH (email@example.com), 1 Joe Menzin, PhD, 1 Kathleen Lang, PhD, 1 Jonathan R. Korn, BA, 1 Peter Neumann, ScD 2 1 Boston Health Economics, Inc., Waltham, MA 2 Tufts Medical Center, Boston, MA
Background <ul><li>The initiation of the Medicare Part D benefit offers an ideal opportunity to study real-world decision-making and the role of adverse selection and other factors in insurance enrollment </li></ul><ul><li>Medical Expenditure Panel Survey (MEPS) data from 2006 are among the few publicly available sources of nationally representative Part D data </li></ul><ul><li>Our objective in this study was to identify predictors of Part D enrollment among individuals with a range of health conditions and insurance designs </li></ul>
Methods Overview <ul><li>Retrospective longitudinal analysis of the MEPS database </li></ul><ul><li>Sample included all individuals in both the 2005 and 2006 MEPS datasets enrolled in Medicare (but not Medicaid) with information available on Part D enrollment </li></ul><ul><ul><li>Excluded dual-eligibles (Medicare + Medicaid enrollees) since they were automatically enrolled in Part D </li></ul></ul><ul><ul><li>Focused on beneficiaries’ insurance coverage as of December 2005, just before enrollment in Part D began </li></ul></ul><ul><li>Descriptive analyses: demographic and socioeconomic characteristics, supplemental insurance, health status, and 2005 medical expenditures </li></ul><ul><li>Multivariate logistic regression used to predict factors associated with the likelihood of enrolling in Part D in 2006 </li></ul>
Data Source <ul><li>MEPS </li></ul><ul><ul><li>Conducted by AHRQ since 1996 </li></ul></ul><ul><ul><li>Annual sample size of ~15,000 households — nationally representative sample of the US civilian noninstitutionalized population </li></ul></ul><ul><ul><li>Each survey panel includes 5 rounds of interviews covering 2 full calendar years </li></ul></ul><ul><li>Expenditures include payments for hospital inpatient stays, ER visits, outpatient visits, office visits, dental visits, prescription medications, home health care, and other medical </li></ul><ul><li>Expenditure data derived from the household component (for amounts paid by respondents) and medical provider component (for amounts paid by third-party payors) </li></ul>
Data Analyses <ul><li>Descriptive analyses performed </li></ul><ul><ul><li>Age, sex, race/ethnicity </li></ul></ul><ul><ul><li>Regional and rural/urban designation </li></ul></ul><ul><ul><li>Marital status, educational attainment, employment status, income category </li></ul></ul><ul><ul><li>Perceived health status, SF-12 scores, number of unique medications </li></ul></ul><ul><ul><li>Supplemental insurance coverage </li></ul></ul><ul><ul><li>Out-of-pocket (OOP) drug expenditures and total OOP expenditures (copayments, co-insurance, deductibles) in 2005 </li></ul></ul><ul><ul><li>Total expenditures (sum of payments by all sources) in 2005 </li></ul></ul><ul><li>Person-level expenditures and weights used </li></ul><ul><li>Premiums excluded from expenditures </li></ul>
Regression Model <ul><li>Multivariate logistic regression analysis performed using the SURVEYLOGISTIC procedure in SAS v9.1 to account for the complex sampling design in MEPS </li></ul><ul><li>Predictors of enrolling in Part D identified using a variety of factors hypothesized to potentially influence Part D enrollment decisions </li></ul><ul><ul><li>Demographics </li></ul></ul><ul><ul><li>Health status </li></ul></ul><ul><ul><li>Income </li></ul></ul><ul><ul><li>Expenditures </li></ul></ul>
Table 1. Demographic and Socioeconomic Characteristics Source: MEPS, 2005-2006
Table 2. Health Status Source: MEPS, 2005-2006. SF-12 scores range from 0-100, with higher scores indicating better health
Figure 1. Supplemental Insurance in 2005 by Enrollment in Part D in 2006 Source: MEPS, 2005-2006
Figure 2. Mean Individual OOP and Total Expenditures in 2005 by Enrollment in Part D in 2006 Source: MEPS, 2005-2006
Figure 3. Multivariate Adjusted Odds of Enrollment in Part D Source: MEPS, 2005-2006
Discussion <ul><li>Concept of adverse selection in insurance markets suggests that, given a choice, healthier people would not enroll in Part D </li></ul><ul><ul><li>Our study: healthy had lower Part D enrollment rates, but perceived health status/SF-12 scores were not significant predictors of Part D enrollment after adjustment </li></ul></ul><ul><li>Concern that less educated and poorer seniors would not be able to make rational enrollment and plan decisions </li></ul><ul><ul><li>Our study: education was a not significant predictor, but low income was positively associated with enrollment </li></ul></ul><ul><li>Extent of existing drug coverage and OOP spending both predicted enrollment decisions </li></ul><ul><li>Less enthusiasm for enrollment among those with 2005 spending in the range of the plan’s coverage gap </li></ul>
Limitations <ul><li>Subject to the usual limitations of retrospective studies and survey research in general </li></ul><ul><li>Unable to analyze whether beneficiaries who did not enroll in Part D had “creditable coverage” </li></ul><ul><ul><li>Type of supplementary insurance served as proxy for this factor </li></ul></ul><ul><ul><ul><li>Few Medigap plans offered drug benefits in 2005* </li></ul></ul></ul><ul><ul><ul><li>Nearly all retiree health plans offered drug benefits in 2005** </li></ul></ul></ul><ul><li>Other, unmeasured factors may have influenced decision to enroll in Part D </li></ul><ul><li>Findings represent first year experiences only </li></ul>*America’s Health Insurance Plans. A Survey of Medigap Enrollment Trends, July 2006. http://www.ahipresearch.org/PDFs/MedigapProductSurvey10-27-06.pdf. **Kaiser Family Foundation. Kaiser/Hewitt 2005 Survey on Retiree Health Benefits, December 2005. http://www.kaiserfamilyfoundation.org/medicare/upload/7439.pdf
Conclusions <ul><li>Based on first-year data, fears of only the sickest beneficiaries enrolling in Part D and employers withdrawing drug benefits to retirees seem to have been unwarranted </li></ul><ul><li>Existing coverage and high prior drug spending drove the decision to enroll in Part D in what appears to have been a rational way </li></ul>