To the Editor:
High drug costs cause some elderly or disabled patients to take less medication than prescribed or forgo basic needs to pay for medicines.1,2,3 The 2006 Medicare Part D drug benefit was intended to increase economic access to medicines.4 Data from 2006 indicated modest nationwide decreases in cost-related medication non-adherence (CRN) and forgoing basic needs following Part D implementation, but no decline in high rates of CRN among the sickest beneficiaries.5 We analyzed more recent data to determine whether the reductions remained stable in 2007.
Methods
The Medicare Current Beneficiary Survey5 (MCBS) is conducted by the Centers for Medicare and Medicaid Services (CMS) to inform and evaluate health policies. In-person interviews collect data on health and medical care from a nationally representative, rotating panel of Medicare enrollees. All community-dwelling respondents from 2004 through 2007 were included (N=14500 [2004], 14701 [2005], 14732 [2006], 14804 [2007]; 29023 unique respondents).
The fall MCBS interview includes questions regarding CRN,2,5 defined as ever (in the current year) skipping or taking smaller doses to make a medicine last longer or not filling a prescription because it was too expensive. Self-reports of ever spending less on food, heat, or other basic needs to afford medicine were also examined.
Prevalence rates of CRN and forgoing basic needs from 2004 to 2007 were calculated for the overall population and in four subgroups: elderly (65+ years) vs nonelderly disabled (<65 years) respondents, each group categorized as 0-2 vs 3 or more morbidities. Odds ratios (ORs) were estimated for 2007 vs 2005 and 2007 vs 2006 using survey-weighted logistic regression models controlling for defined covariates (age group, sex, race, income, general health status, survey participation) and number of morbidities (0-2 vs 3 or more).5 All analyses used Stata version 10 (StataCorp LP, College Station, Texas), 2-sided tests, and statistical significance P<.05. Respondents provided oral informed consent. The study was approved by the Human Subjects Committee of Harvard Pilgrim Health Care.
Results
Unadjusted prevalence rates and the adjusted ORs estimating changes in CRN and forgoing basic needs are shown in the TABLE. Prevalences were consistently higher for nonelderly disabled vs elderly beneficiaries, and for sicker vs healthier beneficiaries. Between 2006 and 2007, there was a very small but statistically significant decrease in CRN for the overall population; in the four subgroups, statistically significant decreases in CRN and forgoing basic needs occurred only for disabled beneficiaries with 3 or more morbidities. Compared with 2005, prior to Part D, the prevalences of CRN and forgoing basic needs in 2007 declined significantly for the overall population and for all four subgroups (ORs between 0.58 and 0.77; P values <.05).
TABLE.
Prevalence Rates from 2004 to 2007 and Changes in Cost-Related Nonadherence and Forgoing Basic Needs Among Community-Dwelling Medicare Beneficiaries
| Subgroup | No. Obs. |
Unadjusted Prevalence, % | 2007 vs 2006 | 2007 vs 2005 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||
| 2004 | 2005 | 2006 | 2007 | OR | 95% CI |
P value |
OR | 95% CI |
P value |
||
| Cost-Related Medication Nonadherence | |||||||||||
|
| |||||||||||
| Overall population | 58,647 | 15.2 | 14.1 | 11.5 | 10.7 | 0.91 | (0.84- 0.99) |
.02 | 0.71 | (0.65-0.79) | < .001 |
| Elderly, 0-2 morbidities | 23,943 | 10.6 | 9.9 | 6.9 | 6.4 | 0.94 | (0.80- 1.10) |
.41 | 0.63 | (0.52-0.75) | < .001 |
| Elderly, 3+ morbidities | 23,814 | 14.8 | 12.9 | 10.4 | 9.9 | 0.96 | (0.85- 1.08) |
.49 | 0.76 | (0.65-0.88) | < .001 |
| Non-elderly Disabled, 0-2 morbidities |
5,301 | 21.5 | 24.5 | 19.3 | 19.4 | 0.98 | (0.74- 1.30) |
.91 | 0.74 | (0.57-0.95) | .02 |
| Non-elderly Disabled, 3+ morbidities |
4,723 | 35.4 | 33.4 | 34.0 | 28.0 | 0.74 | (0.60- 0.92) |
.01 | 0.77 | (0.60- 0.995) |
.046 |
|
| |||||||||||
| Spent Less on Basic Needs | |||||||||||
|
| |||||||||||
| Overall population | 58,457 | 10.6 | 11.1 | 7.6 | 7.8 | 1.03 | (0.92- 1.15) |
.61 | 0.66 | (0.59-0.74) | < .001 |
| Elderly, 0-2 morbidities | 23,867 | 6.2 | 6.8 | 3.5 | 4.0 | 1.18 | (0.95- 1.47) |
.14 | 0.58 | (0.47-0.70) | < .001 |
| Elderly, 3+ morbidities | 23,781 | 10.4 | 10.8 | 6.4 | 7.3 | 1.19 | (1.03- 1.38) |
.02 | 0.66 | (0.55-0.79) | < .001 |
| Non-elderly Disabled, 0-2 morbidities |
5,274 | 16.7 | 19.3 | 14.2 | 13.9 | 0.96 | (0.68- 1.36) |
.82 | 0.69 | (0.51-0.93) | .02 |
| Non-elderly Disabled, 3+ morbidities |
4,717 | 29.5 | 28.4 | 27.9 | 23.3 | 0.78 | (0.61- 0.99) |
.04 | 0.76 | (0.59-0.98) | .04 |
Abbreviations: CI, confidence interval; OR, odds ratio.
ORs were estimated using multivariate logistic regression with survey weights to represent national populations of approximately 37 million to 39 million community-dwelling Medicare beneficiaries per year during this period. Actual numbers of observations vary with item response rates and availability of information on subgroup characteristics. Morbidity categories were described previously.5
Comment
Changes in CRN represent a crucial intermediate step between expanded economic access and potential health improvements. Previously, modest statistically significant reductions in CRN and forgoing basic needs for Medicare beneficiaries following implementation of Part D in 2006 were found, controlling for historical changes. More recent data confirm that these reductions were sustained in 2007. The very small 2006 to 2007 reduction in CRN for the overall population was similar to that estimated between 2004 and 2005,5 prior to Part D. Study limitations include lack of utilization data, likely underestimation of CRN, and the possibility that contemporaneous phenomena could explain the changes observed.5
Subgroup analyses suggest that sicker disabled beneficiaries experienced lagged improvements, apparent only in 2007; such patients may require more time to adapt to administrative changes and realize benefits. Nevertheless, post-Part D reductions in CRN were small in comparison to the persistent disparities in CRN associated with disability and multiple chronic conditions.
ACKNOWLEDGEMENTS
Funding/Support: This study was supported by the National Institute on Aging (NIA) grants # R01AG028745 and # RO1 AG022362, and the Harvard Pilgrim Health Care Foundation. Drs. Ross-Degnan, Briesacher, and Soumerai are investigators in the HMO Research Network Center for Education and Research in Therapeutics, supported by the U.S. Agency for Healthcare Research and Quality (Grant # 2U18HS010391).
Role of the Sponsors: The funding organizations did not participate in the design or conduct of the study, in the collection, analysis or interpretation of the data, or in the preparation, review, or approval of the manuscript.
Footnotes
Author Contributions: Dr Madden had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Soumerai, Madden, Ross-Degnan, Briesacher.
Acquisition of the data: Madden, Soumerai.
Analysis and interpretation of the data: Madden, Soumerai, Graves, Ross-Degnan, Briesacher.
Drafting of the manuscript: Madden.
Critical revision of the manuscript for important intellectual content: Madden, Soumerai, Ross-Degnan, Briesacher, Graves.
Statistical analysis: Graves, Madden.
Obtained funding: Soumerai, Madden, Ross-Degnan, Briesacher.
Administrative, technical, or material support: Graves.
Study supervision: Madden, Soumerai.
Financial Disclosures: Dr Madden reported receiving research support from AstraZeneca. Drs Ross-Degnan and Soumerai and Ms Graves reported receiving research support from Eli Lilly. Dr Briesacher reported prior research support and consulting fees from Novartis.
Additional contributions: Fang Zhang, PhD, of HMS and HPHCI, Alyce S. Adams, PhD, of the Kaiser Permanente Division of Research, Jerry H. Gurwitz, MD, of the Meyers Primary Care Institute and University of Massachusetts Medical School, and Gerald S. Adler, MPhil, of the Centers for Medicare and Medicaid Services (CMS) also contributed to this research. Dana Gelb Safran, ScD, Blue Cross Blue Shield of Massachusetts and Tufts University School of Medicine, provided foundational work on CRN and prior collaborations. Franklin Eppig, JD, and Andrew Shatto of CMS provided consistent support. Robert LeCates, MA, of HMS & HPHCI provided assistance during manuscript preparation. Drs Zhang, Adams, Gurwitz, and Safran and Mr LeCates received partial salary support from the NIA grants mentioned above. Dr Eppig, Mr Adler, and Mr Shatto received no direct compensation.
REFERENCES
- 1.Safran DG, Neuman P, Schoen C, et al. Prescription drug coverage and seniors: Where do things stand on the eve of implementing the new part D benefit? Findings from a 2003 National survey of seniors. Health Aff. 2005 Apr 19; doi: 10.1377/hlthaff.w5.152. (web exclusive): W5-152-66. [DOI] [PubMed] [Google Scholar]
- 2.Soumerai SB, Pierre-Jacques M, Zhang F, et al. Cost-related medication nonadherence among the elderly and the disabled: A national survey one year before the Medicare drug benefit. Arch Intern Med. 2006;166:1829–35. doi: 10.1001/archinte.166.17.1829. [DOI] [PubMed] [Google Scholar]
- 3.Briesacher BA, Gurwitz JH, Soumerai SB. Patients at-risk for cost-related medication nonadherence: a review of the literature. J Gen Intern Med. 2007;22:864–71. doi: 10.1007/s11606-007-0180-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.The Henry J.Kaiser Family Foundation [accessed April 21, 2009];The Medicare Prescription Drug Benefit Fact Sheet. 2009 Mar; Available at: http://www.kff.org/medicare/upload/7044-09.pdf.
- 5.Centers for Medicare and Medicaid Services [accessed April 21, 2009];Appendix A. Sourcebook Series: Health and Health Care of the Medicare Population. Technical Documentation for the Medicare Current Beneficiary Study. 2002 Available at: http://www.cms.hhs.gov/mcbs/downloads/HHC2002appendixA.pdf.