Enrollees who join Medicare Advantage undergo significant turnover in the years following enrollment.
Objectives: To characterize the proportion of Medicare Advantage (MA) enrollees who switched insurers or disenrolled to traditional Medicare (TM) in the years immediately after first choosing to join an MA health plan.
Study Design: Retrospective analysis using 2012-2017 Medicare enrollment data.
Methods: We studied enrollees who joined MA between 2012 and 2016 and identified all enrollees who changed insurers (switched insurance or disenrolled to TM) at least once between the start of enrollment and the end of the study period. We categorized each change as switching insurers or disenrollment to TM, and by whether the previous insurer had exited the market.
Results: Among 6,520,169 new MA enrollees, 15.6% had changed insurance within 1 year after enrollment in MA and 49.2% had changed insurance by 5 years. More enrollees switched insurers rather than disenrolled, and most enrollees who changed insurers did not do so as a result of insurer exits.
Conclusions: New MA enrollees change insurers at a substantial rate when followed across multiple years. These changes may disincentivize insurers from investing in preventive care and chronic disease management and, as shown in several non-MA populations, may lead to discontinuities in care, increased expenditures, and inferior health outcomes.
Am J Manag Care. 2022;28(10):539-542. https://doi.org/10.37765/ajmc.2022.89251
We performed a longitudinal analysis of enrollment turnover among new Medicare Advantage (MA) enrollees. We found the following:
Turnover among enrollees of health plans is a well-studied topic in the commercial,1 Medicaid,1-3 and Affordable Care Act marketplaces.4 Higher rates of insurance switching and disenrollment within these markets have been linked to adverse health outcomes, including delays in seeking medical care,5 lower adherence to medications,2 and increased utilization of higher-cost services such as the emergency department or hospital.2,6 Furthermore, high enrollee turnover may reduce the incentive for health plans to invest in the delivery of certain preventive health services and the management of some chronic medical conditions, as it may take years of continued enrollment before these benefits justify the up-front investment by health plans or other risk-bearing organizations such as accountable care organizations.
Although some research has found increased turnover in Medicare Advantage (MA) patients with high health care needs,7-10 turnover among the general MA population has received less attention. One study that examined rates of plan switching and disenrollment concluded that only a small percentage of MA enrollees switch plans each year.11 Other research has largely focused on the problems that older persons may face when trying to switch plans, such as difficulty interpreting plan information, low awareness of quality ratings, and overall frustration with the plan selection process.12-14 Overall, MA enrollees are often presumed to be “sticky,” tending to stay with the same insurer once enrolled.15
Differences in health outcomes can become magnified over several years, and insurers may account for enrollment changes over multiple years when planning for future expenditures. The cumulative rate of insurance change after a period of enrollment, rather than yearly rates of change, may better characterize the degree of turnover within MA plans. Thus, we performed a longitudinal analysis of MA enrollment outcomes to measure rates of turnover across multiple years.
Using data from the CMS Master Beneficiary Summary File (MBSF),16 we tracked the enrollment status between 2012 and 2017 of new MA enrollees, defined as individuals of any age not enrolled in MA the prior year who then choose to enroll. We focused on these enrollees both to reduce confounding by duration of MA enrollment prior to the study and because recent enrollees comprise a significant portion of the MA population17; we verifiedthe latter with data from the MBSF.
We followed each enrollee over time until either a change in insurance, defined as switching insurers or disenrolling to traditional Medicare (TM), or death occurred. Because insurers, otherwise known as parent organizations, can offer multiple plans to enrollees in the same geographic area, we measured switching among insurers, rather than between plans, to better assess the financial impact of turnover on insurers.
We grouped enrollees into cohorts based on year of enrollment and tracked each cohort’s enrollment outcomes annually. For each period, we calculated the percentage of enrollees who had changed insurance at least once since the start of enrollment. Each insurance change was further subcategorized as involuntary if the original insurer exited the enrollee’s county of residence, voluntary if the insurer did not, or because of a move if the enrollee’s county of residence changed; we then calculated trends for each type. Enrollees who left their original insurer due to death were distinguished from those who switched or disenrolled and were included in the denominator but not the numerator when calculating rates of insurance change.
We excluded enrollees in plans that are not available for individual enrollment, such as employer-sponsored plans, Special Needs Plans, Program of All-Inclusive Care for the Elderly plans, Medicare cost plans, Medicare-Medicaid plans, and demonstration plans, because findings for enrollees in these plans are less applicable to the general Medicare population. We used CMS crosswalk files to identify insurer acquisitions and treated these events as if enrollees remained with their original insurer, as the acquiring insurer would be financially responsible for the acquired enrollees.18 In our supplemental analysis, we stratified by dual-eligibility status and whether the beneficiary was new to Medicare (comparing new Medicare beneficiaries with those who switched from TM). We also compared select demographic characteristics of beneficiaries who changed insurance with those of beneficiaries who did not.
We studied 6,520,169 new MA enrollees who joined between 2012 and 2016, representing approximately 40% of the 16.2 million unique Medicare beneficiaries who were enrolled in individual MA plans between 2012 and 2017. Overall, 15.6% of new MA enrollees had changed insurance 1 or more times by the 1-year mark, 27.7% by 2 years, 37.0% by 3 years, 43.7% by 4 years, and 49.2% by 5 years (Figure 1 and eAppendix [available at ajmc.com]). Rates decreased somewhat across subsequent cohorts, with 1-year rates falling from 18.3% in the 2012 cohort to 14.0% in the 2016 cohort and 3-year rates falling from 40.3% in the 2012 cohort to 33.1% in the 2014 cohort.
When stratified by type of insurance change, voluntary switching was the most common, averaging 7.1% by 1 year and 22.9% by 5 years of enrollment, although rates declined somewhat across later cohorts (Figure 2 and eAppendix). The rate of voluntary disenrollment to TM was approximately half that of voluntary switching and remained stable across successive cohorts. In contrast, the rate of involuntary switching decreased sharply between the 2012 and 2014 cohorts before stabilizing thereafter, whereas involuntary disenrollment to TM saw a smaller decrease across the same time frame. When accounting for the size of each subgroup, the decrease in the overall rate of insurance change appears largely driven by decreases in voluntary and involuntary switching prior to the 2015 cohort.
Dual-eligible enrollees changed insurance at a higher rate than that of the general population, although restricting the analysis to non–dual-eligible enrollees, new Medicare beneficiaries, or enrollees who switched from TM changed our results by just 1% to 2% each year (full results in eAppendix). Enrollees who changed insurance, especially those who voluntarily disenrolled, were more likely to have switched from TM when joining MA, to be dual-eligible, and to have more chronic conditions on average (eAppendix).
We show that after initial enrollment, MA enrollees continue to change insurance over the next several years, and that insurance changes are driven predominantly by voluntary choices. Relative to other systems, one may naturally expect lower rates of turnover within MA given that enrollment is not affected by changes in employment, as with commercial and ACA plans, or income, as with Medicaid, and because most MA enrollees can only change insurance only during open enrollment periods. Remarkably, the rates we report are comparable with rates found in certain non-Medicare populations; for example, 26% of enrollees in a mixed commercial and Medicaid population switched insurers within 2 years, and 13.7% of Medicaid enrollees in expansion states changed or lost coverage within 1 year.1,3
In an earlier report, Jacobson et al followed MA enrollees across 1-year intervals and found that only a small proportion of enrollees change plans each year.11 Although our conclusion may appear different because of Jacobson et al measuring plan-level insurance change, focusing on voluntary switching, and including existing MA enrollees, our 1-year switching rates are similar to theirs after accounting for the above (eAppendix). The main reason for the difference appears to be the longer time frame we used, as changes that are perceived as small on a year-to-year basis can add up significantly over time. Our results complement those of prior research by highlighting that the cumulative effect of turnover may have significant downstream effects on both insurers and enrollees.
A key underlying principle of managed care is that investments in preventive health and improved management of chronic conditions will lead to better long-term patient outcomes and, potentially, lower utilization over time. Insurers, however, are more likely to make such investments if they stand to benefit from the future decreased expenditures. Because Medicare provides coverage for almost all individuals 65 years and older, the Medicare program itself should benefit from efforts to prevent future disease. This does not mean, however, that MA plans, which patients can elect to join and leave on a yearly basis, will benefit in the same way. Our data suggest, in fact, that the turnover in MA enrollment may reduce incentives for MA insurers to invest in better management of chronic medical conditions.
One may assume that if enrollees switch insurers frequently, a significant portion of those who switched would eventually reenroll with their original insurer, thus obviating the financial disincentive toward future investment. However, insurers who do not invest in these services may be able to “free ride” from those who do, creating an incentive against future investment. This effect has been observed empirically in other health insurance markets where reenrollment could also occur.19 Furthermore, discontinuities in enrollment could make the implementation of programs to improve prevention or delay disease progression more difficult—for example, an intervention for enrollees with poorly controlled diabetes may be hindered by spotty data on the frequency of hemoglobin A1c checks—and potentially not as cost-effective.
High turnover rates may also lead to inferior health outcomes among MA enrollees, as has been observed in other insured populations. Discontinuities in medical care have been shown to increase older Americans’ usage of emergency care and hospital length of stay,20 and in non-Medicare populations changes in insurance have also been linked with delays in seeking care, skipping doses of medications, and overall declines in health.1,2 Some enrollees may change insurance to keep access to an existing provider or medication, but a switch may still introduce discontinuities of care elsewhere in their care. Although these effects have been studied predominantly in non-MA populations, the negative effects of high turnover could be magnified in MA as both medical complexity and health care utilization generally increase with age.
Our work also highlights several areas for further research. Our data, like those of prior work,7-10 suggest that high-need enrollees disenroll at higher rates. It would be important to both understand their reasons for disenrollment and assess whether any barriers, such as coverage gaps in Medigap plans for preexisting conditions, affect subsequent outcomes. Trends in insurance change could also be of interest to policy makers. For example, the decrease in involuntary switching and disenrollment over successive cohorts may reflect the greater financial attractiveness for insurers of MA compared with other insurance markets,21 whereas the decrease in voluntary switching could be suggestive of increased satisfaction among MA enrollees. We caution that the policy implications of these findings are not clear without further investigation.
Our study has several limitations. Because we focused on new enrollees, our results may overestimate the overall rate of insurance change within MA as older enrollees tend to change insurance at lower rates, although our study population did include around 40% of all MA enrollees in individual plans. Rates at 4 and 5 years after initial enrollment may become lower with more follow-up data given that there was less insurance change in the 2015 and 2016 cohorts. Because we did not track enrollee outcomes after an insurance change, we do not know what fraction of enrollees eventually reenroll with their original insurer, although these enrollees are still exposed to multiple insurance transitions as described above. Our definition of new enrollees also does allow those who were previously enrolled in MA but disenrolled for multiple years before rejoining to be included. Finally, although we attempted to estimate the fraction of insurance change due to insurer exits, we note that enrollees labeled with an involuntary change may have had other reasons for switching insurance or disenrolling.
In this study of longitudinal turnover in MA, we show that insurance change among new MA enrollees may be more substantial than previously portrayed, with rates comparable with those of commercial and Medicaid populations. Consequently, turnover in MA may lead to inferior health outcomes through disjointed care and decreased investment in improved care delivery. Future research could help assess the extent to which turnover affects the care that MA enrollees receive.
The authors would like to thank Jeffrey Souza for help with data analysis.
Author Affiliations: Department of Medicine, Beth Israel Deaconess Medical Center (JD, BEL), Boston, MA; Department of Health Care Policy, Harvard Medical School (AMZ, BEL), Boston, MA; Institute for Healthcare Policy and Innovation, Division of General Medicine, Department of Health Management and Policy in the School of Public Health, and Gerald R. Ford School of Public Policy, University of Michigan (JZA), Ann Arbor, MI.
Source of Funding: Supported by a grant from the National Institute on Aging (P01 AG032952).
Author Disclosures: Dr Ayanian is employed as a practicing physician by the University of Michigan and is an unpaid board member of Physicians Health Plan representing Michigan Medicine. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (JD, AMZ, JZA, BEL); acquisition of data (JD, BEL); analysis and interpretation of data (JD, AMZ, JZA, BEL); drafting of the manuscript (JD); critical revision of the manuscript for important intellectual content (JD, AMZ, JZA, BEL); statistical analysis (AMZ); obtaining funding (BEL); and supervision (BEL).
Address Correspondence to: Bruce E. Landon, MD, MBA, Department of Health Care Policy, Harvard Medical School, 180 Longwood Ave, Boston, MA 02215. Email: firstname.lastname@example.org.
1. Barnett ML, Song Z, Rose S, Bitton A, Chernew ME, Landon BE. Insurance transitions and changes in physician and emergency department utilization: an observational study. J Gen Intern Med. 2017;32(10):1146-1155. doi:10.1007/s11606-017-4072-4
2. Sommers BD, Gourevitch R, Maylone B, Blendon RJ, Epstein AM. Insurance churning rates for low-income adults under health reform: lower than expected but still harmful for many. Health Aff (Millwood). 2016;35(10):1816-1824. doi:10.1377/hlthaff.2016.0455
3. Goldman AL, Sommers BD. Among low-income adults enrolled in Medicaid, churning decreased after the Affordable Care Act. Health Aff (Millwood). 2020;39(1):85-93. doi:10.1377/hlthaff.2019.00378
4. Gordon SH, Sommers BD, Wilson IB, Galarraga O, Trivedi AN. Risk factors for early disenrollment from Colorado’s Affordable Care Act marketplace. Med Care. 2019;57(1):49-53. doi:10.1097/MLR.0000000000001020
5. Burstin HR, Swartz K, O’Neil AC, Orav EJ, Brennan TA. The effect of change of health insurance on access to care. Inquiry. 1998;35(4):389-397.
6. Banerjee R, Ziegenfuss JY, Shah ND. Impact of discontinuity in health insurance on resource utilization. BMC Health Serv Res. 2010;10:195. doi:10.1186/1472-6963-10-195
7. Rahman M, Keohane L, Trivedi AN, Mor V. High-cost patients had substantial rates of leaving Medicare Advantage and joining traditional Medicare. Health Aff (Millwood). 2015;34(10):1675-1681. doi:10.1377/hlthaff.2015.0272
8. Li Q, Trivedi AN, Galarraga O, Chernew ME, Weiner DE, Mor V. Medicare Advantage ratings and voluntary disenrollment among patients with end-stage renal disease. Health Aff (Millwood). 2018;37(1):70-77. doi:10.1377/hlthaff.2017.0974
9. Meyers DJ, Belanger E, Joyce N, McHugh J, Rahman M, Mor V. Analysis of drivers of disenrollment and plan switching among Medicare Advantage beneficiaries. JAMA Intern Med. 2019;179(4):524-532. doi:10.1001/jamainternmed.2018.7639
10. Meyers DJ, Rahman M, Rivera-Hernandez M, Trivedi AN, Mor V. Plan switching among Medicare Advantage beneficiaries with Alzheimer’s disease and other dementias. Alzheimers Dement (N Y). 2021;7(1):e12150. doi:10.1002/trc2.12150
11. Jacobson G, Neuman T, Damico A. Medicare Advantage plan switching: exception or norm? Kaiser Family Foundation. September 20, 2016. Accessed June 29, 2021. https://www.kff.org/medicare/issue-brief/medicare-advantage-plan-switching-exception-or-norm/
12. Goldstein E, Fyock J. Reporting of CAHPS quality information to Medicare beneficiaries. Health Serv Res. 2001;36(3):477-488.
13. Biles B, Dallek G, Nicholas LH. Medicare advantage: déjà vu all over again? Health Aff (Millwood). 2004;23(suppl 1):W4-586-W4-597. doi:10.1377/hlthaff.w4.586
14. Jacobson G, Swoope C, Perry M, Slosar MC. How are seniors choosing and changing health insurance plans? Kaiser Family Foundation. May 13, 2014. Accessed June 29, 2021. https://www.kff.org/medicare/report/how-are-seniors-choosing-and-changing-health-insurance-plans/
15. Neuman P, Jacobson GA. Medicare Advantage checkup. N Engl J Med. 2018;379(22):2163-2172. doi:10.1056/NEJMhpr1804089
16. Master Beneficiary Summary File (MBSF) base. Research Data Assistance Center. Accessed June 29, 2021. https://www.resdac.org/cms-data/files/mbsf-base
17. The Medicare Advantage program: status report. In: Medicare Payment Advisory Commission. Report to the Congress: Medicare Payment Policy. Medicare Payment Advisory Commission; 2020. Accessed June 29, 2021. https://www.medpac.gov/wp-content/uploads/import_data/scrape_files/docs/default-source/reports/mar20_medpac_ch13_sec.pdf
18. Plan crosswalks. CMS. Accessed June 29, 2021. https://www.cms.gov/Research-Statistics-Data-
19. Herring B. Suboptimal provision of preventive healthcare due to expected enrollee turnover among private insurers. Health Econ. 2010;19(4):438-448. doi:10.1002/hec.1484
20. Wasson JH, Sauvigne AE, Mogielnicki RP, et al. Continuity of outpatient medical care in elderly men. a randomized trial. JAMA. 1984;252(17):2413-2417. doi:10.1001/jama.1984.03350170015011
21. Jacobson G, Fehr R, Cox C, Neuman T. Financial performance of Medicare Advantage, individual, and group health insurance markets. Kaiser Family Foundation. August 5, 2019. Accessed November 28, 2021. https://www.kff.org/report-section/financial-performance-of-medicare-advantage-individual-and-group-health-insurance-markets-issue-brief/