Plan report example

Attachment IX:
Redacted Plan Report Sample

2017 Medicare Advantage Health and Drug Plan
Disenrollment Reasons Survey Results
Report for:

(

)

Issued August 2018
By the Centers for Medicare & Medicaid Services

Contents
AttachmentIX:RedactedPlan Report Sample .............................................................................. 1
2017 Medicare Advantage Health and Drug Plan Disenrollment Reasons Survey Results.... 2
Contents.......................................................................................................................... 1
I.

OVERVIEW................................................................................................................. 3
How This Report Is Organized ....................................................................................... 3
How Scores Are Compared ........................................................................................... 3
How to use this report.................................................................................................. 4

II.

SUMMARY RESULTS ............................................................................................... 4
Disenrollment Rate....................................................................................................... 4
Table 1.1. Rates of Voluntary Disenrollment .............................................................. 4
Summary of Characteristics of Enrollees and their Disenrollment Rates ......................... 5
Table 1.2. .................................................................................................................... 5
Reasons for Disenrollment............................................................................................ 6
Table 1.3. Composite Measures of Reasons for Disenrollment .................................. 6
Table 1.4. Individual Survey Questions ....................................................................... 7

III.

DETAILED RESULTS FOR COMPOSITE MEASURES ..................................................... 8

Problems with Coverage of Doctors and Hospitals......................................................... 8
Problems with Coverage of Doctors and Hospitals ..................................................... 8
Financial Reasons for Disenrollment ............................................................................. 8
Financial Re asons for Disenrollment........................................................................... 9
Problems Getting Needed Care, Coverage, and Cost Information................................... 9
Problems Getting Needed Care, Cove rage, and Cost Information.............................10
Problems Getting Information and Help from the Plan .................................................10
Problems Getting Information and Help from the Plan .............................................11
Problems with Prescription Drug Benefits and Coverage ..............................................11
Problems with Prescription Drug Benefits and Cove rage ..........................................12
IV.

BACKGROUND AND METHODOLOGY .....................................................................12
1

Methodology ..............................................................................................................13
Table 1.5. Medicare Advantage Health and Drug Plan Disenrollment Reasons Survey
Composite Measures .................................................................................................14
Table 1.6. Individual Survey Questions Not Included in a Composite Measure* .......15
Reporting of Composite Means on the Medicare Plan Finder........................................15
Table 1.7. Composite Measure Labels and Mean Scores as Presented in This Report
vs. the Medi care Plan Finder .....................................................................................16
Sample Selection and Eligibility Criteria........................................................................16
Survey Implementation ...............................................................................................17
Sample Disposition......................................................................................................17
Table 1.8. 2017 Medicare Advantage and Prescription Drug Plan Disenrollment
Reasons Survey, MA Sample Disposition ...................................................................18
Weighting and Case-Mix Adjustment ...........................................................................18
Assessing Reliability of Scores......................................................................................20
Significance Testing .....................................................................................................21
Comparison of Reasons for Disenrollment: January-November versus December
Disenrollees ................................................................................................................22
Table 1.9. National Average Scores on Composite Measures and Their Constituent
Questions: January-November Disenrollees vs. Dece mber Disenrollees ...................23
Table 1.10. National Average Scores on Single Questions (not in a composite):
January-Novembe r Disenrollees vs. Dece mber Disenrollees .....................................23
State or Regional Comparisons ....................................................................................24
Contact Information ....................................................................................................24

2

I. OVERVIEW
This report contains results for your contract
from the 2017 Medicare Advantage and
Prescription Drug Plan Disenrollment Reasons Survey, a Centers for Medicare & Medicaid
Services-sponsored survey that assesses the reasons for disenrollment among Medicare
beneficiaries who have voluntarily disenrolled from their Part C and Part D contracts. The survey
period covers disenrollment that occurred between January and December 2017. Although
beneficiaries provide ratings of their “plans,” the unit of analysis is the health and/or
prescription drug plan contract, not a health and/or prescription drug plan (i.e., Plan Benefit
Package or PBP). These contract-specific reports are sent to compliance officers on an annual
basis.
When considered in conjunction with disenrollment rates, beneficiaries' responses to the survey
provide information about the quality of a contract as it is experienced by beneficiaries who
have chosen to disenroll from the contract. CMS is sharing detailed results from the 2017 survey
with contracts to facilitate quality improvement efforts. CMS displays your contract’s
disenrollment rate and scores on the composite (summary) disenrollment reason measures on
Medicare Plan Finder (https://www.medicare.gov/find-a-plan) for use by consumers when
selecting contracts (i.e., health and prescription drug plans).

How This Report Is Organized
Section II of this report provides summary results on your contract’s disenrollment rate, a
comparison of your contract’s beneficiary characteristics to the characteristics of Medicare
Advantage (MA) beneficiaries nationally, and reasons beneficiaries cited for disenrollment, both
composite (summary) measures and individual survey questions. Section III presents detailed
results on composite measures of reasons for disenrollment, showing both national and state
benchmarks to enable comparisons to your contract’s results. Section IV provides information
about the survey and its contents and describes sample selection and other methodological
topics. A separate Excel file distributed with this report contains frequency tables that display
unadjusted (i.e., not adjusted for case-mix) responses to all survey questions.

How Scores Are Compared
Your contract's disenrollment rate (Table 1.1), mean scores on the composite measures of
reasons for disenrollment (Table 1.3), and results on individual survey questions underlying the
composite measures (Table 1.4) are each compared to the national average for all MA plans. In
both Tables 1.3 and 1.4, the results for the composite and individual reasons for disenrollment
are ordered from most frequently cited to least frequently cited based on the national average.
Each composite measure score is tested against the national average for that composite
measure, and a note in the last column of Table 1.3 indicates whether the difference is
3

statistically significant. Table 1.4 provides results, including comparisons to the national average,
for individual survey questions about reasons for disenrollment. For comparison, Tables 1.3 and
1.4 also show a state or regional average; however, because of a lack of statistical power,
statistical tests of the difference between your contract and the state or regional average were
not performed.

How to use this report
MA-PD plans can use the information in this report to identify areas of strong performance and
opportunities for improvement. Comparing the reasons beneficiaries give for voluntarily
disenrolling from your contract with the reasons beneficiaries give for leaving MA contracts
nationally may provide insights about your contract's strengths and weaknesses.

II.

SUMMARY RESULTS

Disenrollment Rate
Table 1.1 shows the rate of voluntary disenrollment from your contract for calendar year 2017
(January 2017 to December 2017) and 2016 (January 2016 to December 2016). These
disenrollment rates are calculated as a percentage of your contract's total enrollment each year
and come from beneficiary-level disenrollment files maintained by the Centers for Medicare &
Medicaid Services. The table also shows the national average rate of voluntary disenrollment for
MA contracts in 2017 and whether the difference between your contract's rate of voluntary
disenrollment in 2017 and the national average rate of disenrollment was statistically (p < 0.05)
and practically (at least one percentage point) significant.
Voluntary disenrollment refers to a beneficiary either dropping coverage entirely or switching to
another contract for coverage. Excluded from this calculation are beneficiaries who involuntarily
disenrolled from your contract because they were no longer eligible for coverage, moved out of
your contract's service area, switched benefits packages within your contract, were involuntarily
reassigned or passi1vely enrolled in a Medicare-Medicaid Plan, or died.
Table 1.1. Rates of Voluntary Disenrollment
Your Contract

Your Contract

(

(

), 2016

), 2017

National
Average, 2017

Significantly Different from
National Average in 2017?

12%
The disenrollment rates shown in Table 1.1 are the same disenrollment rates used in the Medicare Star
Ratings. The voluntary disenrollment rate is calculated according to the Medicare 2019 Part C & D Star
Rating Technical Notes for “Members Choosing to Leave the Plan.” Rates shown for 2017 are based on
the first plan preview of 2019 Star Ratings data. These may differ from the final rates, found on
Medicare Plan Finder for 2019 Part C & D Star Ratings. If a disenrollment rate is not shown for your
contract, it is because your contract had fewer than 1,000 enrollees in the year for which the rate is not
shown. Disenrollment rates are not computed for contracts with fewer than 1,000 enrollees.

4

Beneficiaries may disenroll from their Medicare health (MA-Only or MA-PD) or prescription drug
plans (PDPs) at different times of the year. The majority of beneficiaries who disenroll do so in
the month of December as part of the annual Medicare Open Enrollment Period (OEP). 1
Beneficiaries may switch plans outside the Open Enrollment Period under special circumstances,
such as if they move out of the plan’s service area, are dually eligible for Medicare and Medicaid,
qualify for the Low Income Subsidy (LIS), qualify for membership in a Special Needs Plan (SNP),
or enroll in a plan with a 5-star rating. In 2017, the percent of all disenrollees from MA plans
who disenrolled in December was 55.0% whereas the percent of disenrollees from your contract
who disenrolled in December was
. Table 1.9 (shown in Section IV of the report) contains
information to help you understand how beneficiaries who disenroll in December, nationally,
compare with those who disenroll during other times of the year in terms of their reasons for
disenrollment.

Summary of Characteristics of Enrollees and their Disenrollment Rates
The first two columns in Table 1.2 compare the characteristics of beneficiaries enrolled in your
contract to the characteristics of MA beneficiaries nationally. The last two columns of Table 1.2
show the disenrollment rate for beneficiaries in your contract, overall and by certain
characteristics, and how the disenrollment rates for your contract compare with rates observed
nationally among beneficiaries with the same characteristics.
Table 1.2.
Percent of
Enrollment
Characteristic
ALL Beneficiaries

National
100%

Percent of
Enrollment
Your Contract
(

)

Disenrollme nt
Rate
National
12%

Dual-eligible

33%

13%

Eligible for Low-Income
Subsidy (LIS)/Not Dual-eligible
Not LIS-eligible/Not Dualeligible
Senior (age 65)

4%

12%

63%

9%

78%

10%

Non-senior (age <65)

22%

11%

Disenrollme nt
Rate
Your Contract
(

)

Note: The voluntary disenrollment rate for "ALL Beneficiaries" (row 1) is calculated according to the Medicare 2019 Part C & D Star
Rating Technical Notes for “Members Choosing to Leave the Plan.” The rates shown are based on the first plan preview of 2019 Star
Ratings data.These may differ from the final rates, found on Medicare Plan Finder for2019 PartCand D Star Ratings. For the voluntary
disenrollment rates for the subgroup categories (rows 2-6), we were unable to apply all of the same exclusions applied to the
disenrollment rate for all beneficiaries; therefore, these rates may appear slightly higher or lower than the rates for “ALL
Beneficiaries.” If a disenrollment rate is not shown for your contract, it is because your contract had fewer than 1,000 enrollees in
the year for which the rate is not shown. Disenrollment rates are not computed for contracts with fewer than 1,000 enrollees.
1

The Medicare Open Enrollment Period runs from October 15th through December 7th annually, but disenrollments that
occur w ithin the Open Enrollment Period are not effective until December and show up as December disenrollees.

5

Reasons for Disenrollment
Responses to individual survey questions were combined to form five composite (summary)
measures of reasons for disenrollment. For each composite measure, Table 1.3 shows your
contract’s mean score on a 0-100 scale, the national average for all MA contracts, a state or
regional average, and whether your contract’s mean was significantly different from the national
average. The mean score for your contract is the average percentage of reasons endorsed in a
composite multiplied by 100. See Section IV, Background and Methodology, of this report (p.16)
for an example of how composite means are calculated. A lower mean score on a measure
indicates that fewer disenrollees from your contract cited the reasons for disenrollment included
in the composite as a cause of their disenrollment.
Table 1.4 displays your contract’s results on the individual survey questions that are included in
each composite, as well as other individual survey questions about reasons for disenrollment
that are not included in the composite measures. Your contract’s score on an individual survey
question is simply the percentage of survey respondents who endorsed the reason as a cause of
their disenrollment.
In both Tables 1.3 and 1.4, the composite and individual reasons for disenrollment results are
ordered from most frequently cited to at least frequently cited based on the national average.
Table 1.3. Composite Measures of Reasons for Disenrollment
Your
Contract
Reasons for Disenrollme nt
Problems with Coverage of Doctors and Hospitals

(

)

National
Average
27.4

State
Average:
25.6

Financial Reasons

24.3

26.6

Problems Getting Needed Care, Coverage, and Cost
Information
Problems Getting Information and Help from the
Plan
Problems with Prescription Drug Benefits and
coverage

18.3

22.4

12.8

14.8

12.2

12.8

Significantly
Different from the
National Average?

* In previous reports, this measure was labeled “Problems Getting Information about Prescription Drugs.”
Note: Scores in bold have adequate reliability (0.70 or higher). Scores that are not in bold have low reliability (between 0.60 and
0.70). Scores in brackets have very low reliability (below 0.60) but are statistically significantly different from the national
average and are therefore reported. N/A means too few disenrollees answered the questions that make up the composite to
permit reporting. For information on how we tested for statistical significance, assessed reliability, and adjusted for case-mix,
see Section IV for this report, pp. 18-22.

6

Table 1.4. Individual Survey Questions

Reasons for Disenrollme nt
Problems with Coverage of Doctors and Hospitals
Preferred provider not in plan

Your
Contract
(
)

National
Average

State
Average:

31.6%

29.2%

Preferred clinic or hospital not covered by plan

23.2%

22.0%

Financial Reason
Found a plan that costs less

40.3%

43.9%

Prescription co-payment went up

20.5%

21.9%

Monthly premium went up

19.4%

21.5%

Could no longer afford plan

17.2%

19.1%

Problems Getting Needed Care, Coverage, and Cost
Information
Problems getting needed care, tests, or treatment

23.2%

28.1%

23.2%

28.0%

14.3%

17.0%

12.3%

16.4%

Frustration with approval process for care, tests, or
treatment
Hard to get information about coverage and cost of
health services
Problems getting claims paid
Problems Getting Information and Help from the Plan*
Could not get information or help needed from the plan

20.3%

24.8%

Unhappy with how the plan handled a question or
complaint
Hard to get information about coverage and cost of
prescription drugs
Customer service not courteous or respectful

18.8%

21.9%

10.7%

12.0%

7.6%

8.2%

Did not know whom to contact about filling a
prescription
Problems with Prescription Drug Benefits and Coverage
Frustrating approval process for off-formulary
medications
Problems getting prescribed medication

6.6%

7.1%

13.5%

15.1%

12.7%

12.8%

Plan refused to pay for a prescribed medication

12.5%

14.6%

Change in drug formulary

11.3%

11.2%

Difficult to get brand name medications

10.9%

10.0%

49.1%

49.7%

34.7%

36.1%

Family member or friend recommended another plan

31.3%

30.5%

Saw commercial or advertisement for another plan
that looked better
Co-payment for doctor visit went up

23.8%

23.9%

19.4%

20.4%

Found a plan with a higher Medicare Star rating

15.6%

16.1%

Low Medicare Star rating

5.0%

3.1%

Individual Survey Questions (not in a composite)
Another plan offered better benefits or coverage of
health services
Another plan better met prescription needs

Significantly
Different from the
National Average?

* In previous reports, this measure was labeled “Problems Getting Information about Prescription Drugs.”
Note: Scores in bold have adequate reliability (0.70 or higher). Scores that are not in bold have low reliability (between 0.60
and 0.70). Scores in brackets have very low reliability (below 0.60) but are statistically significantly different from the national
average and are therefore reported. N/A means too few disenrollees answered the question to permit reporting. For
information on how we tested for statistical significance and adjusted for case-mix, see Section IV of this report, pp. 18-22.

7

III.

DETAILED RESULTS FOR COMPOSITE MEASURES

Problems with Coverage of Doctors and Hospitals
The figure below shows how your contract performed on the measure "Problems with Coverage
of Doctors and Hospitals," a composite of survey questions 35 and 36. Each of these questions
asked about a reason for disenrollment that was related to the coverage of doctors and
hospitals by the plan. The figure shows the number of disenrollees from your contract who
answered at least one of these questions and the percentage of those disenrollees who
endorsed 0, 1, or both of the reasons as a cause of their disenrollment. The figure also shows
your contract's mean on the composite (the average percentage of reasons endorsed in the
composite multiplied by 100) and whether the mean was higher or lower than the national
average for all MA contracts. A lower mean indicates that problems with coverage of doctors
and hospitals were cited less frequently by disenrollees from your contract. If the mean for your
contract appears in bold, it signifies that the mean has adequate reliability (0.70 or above in a 0
to 1.0 range).
Means not appearing in bold have low reliability (between 0.60 and 0.70). Means appearing in
brackets have very low reliability (below 0.60) but are statistically significantly different from the
national mean and are therefore reported. N/A signifies that too few disenrollees answered the
question to permit reporting. Results for the individual survey questions that this composite
measure comprises appear in Table 1.4.
Problems with Coverage of Doctors and Hospitals
*= significantly different from
the national average
Mean
National Distribution
State Distribution:

N=27,708

27.4

N=862

25.6

Note: N/A means too few disenrollees answered the questions to permit reporting. Percentages may not add to 100 due to
rounding. The mean score is calculated as the average percentage of reasons endorsed in this composite (0 to 100 scale).
Contract means appearing in bold have adequate reliability (0.70 or above). Contract means not appearing in bold have low
reliability (0.60-0.70). Contract means appearing in brackets have very low reliability (below 0.60) but are statistically significantly
different from the national mean. For information on how we tested for statistical significance, assessed reliability, and adjusted
for case-mix, see Section IV of this report, pp. 18-22.

Financial Reasons for Disenrollment
This figure below shows how your contract performed on the measure “Financial Reasons for
Disenrollment,” a composite of survey questions 20, 22, 24, and 25. Each of these questions

8

asked about a reason for disenrollment that was related to the cost or affordability of services.
The figure shows the number of disenrollees from your contract who answered at least one of
these questions and the percentage of those disenrollees who endorsed 0, 1, or 2 or more of the
reasons as a cause of their disenrollement. The figure also shows your contract’s mean on the
composite (the average percentage of reasons endorsed in the composite multiplied by 100) and
whether the mean was higher or lower than the national average for all MA contracts. A lower
mean indicates that financial reasons were endorsed less frequently by disenrollees from your
contract. If the mean for your contract appears in bold, it signifies that the mean has adequate
reliability (0.70 or above in a 0 to 1.0 range).
Mean scores not appearing in bold have low reliability (between 0.60 and 0.70). Means
appearing in brackets have very low reliability (below 0.60) but are statistically significantly
different from the national mean and are therefore reported. N/A signifies that too few
disenrollees answered the questions to permit reporting. Results for the individual survey
questions that this composite measure comprises appear in Table 1.4.
Financial Reasons for Disenrollment
*= significantly different from
the national average
Mean
National Distribution

State Distribution:

N=27,715

24.3

N=862

26.6

Note: N/A means too few disenrollees answered the questions to permit reporting. Percentages may not add to 100 due to
rounding. The mean score is calculated as the average percentage of reasons endorsed in this composite (0 to 100 scale).
Contract means appearing in bold have adequate reliability (0.70 or above). Contract means not appearing in bold have low
reliability (0.60-0.70) Contract means appearing in brackets have very low reliability (below 0.60) but are statistically significantly
different from the national mean. For information on how we tested for statistical significance, assessed reliability, and adjusted
for case-mix, see Section IV of this report, pp. 18-22.

Problems Getting Needed Care, Coverage, and Cost Information
The figure below shows how your contract performed on the measure "Problems Getting
Needed Care, Coverage, and Cost Information," a composite of survey questions 32, 33, 34, and
37. Each of these questions asked about a reason for disenrollment that was related to the
beneficiary's experiences with getting needed health care services and cost information and
getting claims paid for these services. The figure shows the number of disenrollees from your
contract who answered at least one of these questions and the percentage of those disenrollees
9

who endorsed 0, 1, or 2 or more of the reasons as a cause of their disenrollment. The figure also
shows your contract's mean on the composite (the average percentage of reasons endorsed in
the composite multiplied by 100) and whether the mean was higher or lower than the national
average for all MA contracts. A lower mean indicates that problems getting needed care,
coverage, and cost information were cited less frequently by disenrollees from your contract. If
the mean for your contract appears in bold, it signifies that the mean has adequate reliability
(0.70 or above in a 0 to 1.0 range). Means not appearing in bold have low reliability (between
0.60 and 0.70). Means appearing in brackets have very low reliability (below 0.60) but are
statistically significantly different from the national mean and are therefore reported. N/A
signifies too few disenrollees answered the question to permit report ing. Results for the
individual survey questions that this composite measure comprises appear in Table 1.4.
Problems Getting Needed Care, Coverage, and Cost Information
*= significantly different from
the national average
Mean
National Distribution

State Distribution:

N=27,714

18.3

N=862

22.4

Note: N/A means too few disenrollees answered the questions to permit reporting. Percentages may not add to 100 due to
rounding. The mean score is calculated as the average percentage of reasons endorsed in this composite (0 to 100 scale).
Contract means appearing in bold have adequate reliability (0.70 or above). Contract means not appearing in bold have low
reliability (0.60-0.70). Contract means appearing in brackets have very low reliability (below 0.60) but are statistically significantly
different from the national mean. For information on how we tested for statistical significance, assessed reliability, and adjusted
for case-mix, see Section IV of this report, pp. 18-22.

Problems Getting Information and Help from the Plan
The figure below shows how your contract performed on the measure "Problems getting
Information and Help from the Plan," a composite of survey questions 30, 31, 38, 39, and 40.
Each of these questions asked about a reason for disenrollment that was related to the
beneficiary's experiences with getting information and help from the plan. In previous reports,
this composite was labeled "Problems Getting Information about Prescription Drugs." Although
the label has changed, the questions that make up the composite remain the same. The figure
shows the number of disenrollees from your contract who answered at least one of these
questions and the percentage of those disenrollees who endorsed 0, 1, or 2 or more of the
reasons as a cause of their disenrollment. The figure also shows your contract's mean on the
composite (the average percentage of reasons endorsed in the composite multiplied by 100) and
10

whether the mean was higher or lower than the national average for all MA contracts. A lower
mean indicates that problems getting information and help from the plan were cited less
frequently by disenrollees from your contract. If the mean for your contract appears in bold, it
signifies that the mean has adequate reliability (0.70 or above in a 0 to 1.0 range). Means not
appearing in bold have low reliability (between 0.60 and 0.70). Means appearing in brackets
have very low reliability (below 0.60) but are statistically significantly different from the national
mean and are therefore reported. N/A signifies that too few disenrollees answered the question
to permit reporting. Results for the individual survey questions that this composite measure
comprises appear in Table 1.4.
Problems Getting Information and Help from the Plan
*= significantly different from
the national average
Mean
National Distribution

State Distribution:

N=27,423

12.8

N=862

14.8

Note: N/A means too few disenrollees answered the questions to permit reporting. Percentages may not add to 100 due to
rounding. The mean score is calculated as the average percentage of reasons endorsed in this composite (0 to 100 scale).
Contract means appearing in bold have adequate reliability (0.70 or above). Contract means appearing in brackets have very low
reliability (below 0.60) but are statistically significantly different from the national mean. For information on how we tested for
statistical significance, assessed reliability, and adjusted for case-mix, see Section IV of this report, pp. 18-22.

Problems with Prescription Drug Benefits and Coverage
The figure below shows how your contract performed on the measure "Problems with
Prescription Drug Benefits and Coverage," a composite of survey questions 23, 26, 27, 28, and
29. Each of these questions asked about a reason for disenrollment that was related to
prescription drug benefits and coverage. The figure shows the number of disenrollees from your
contract who answered at least one of these questions and the percentage of those disenrollees
who endorsed 0, 1, or 2 or more of the reasons as a cause of their disenrollment. The figure also
shows your contract's mean on the composite (the average percentage of reasons endorsed in
the composite multiplied by 100) and whether the mean was higher or lower than the national
average for all MA contracts. A lower mean indicates that problems with prescription drug
benefits and coverage were cited less frequently by disenrollees from your contract. If the mean
for your contract appears in bold, it signifies that the mean has adequate reliability (0.70 or
above in a 0 to 1.0 range). Means not appearing in bold have low reliability (between 0.60 and
0.70).
11

Means appearing brackets have very low reliability (be low 0.60) but are statistically significantly
different from the national mean and are therefore reported. N/A signifies that too few
disenrollees answered the question to permit report ing. Result s for the individual survey
questions that this composite measure comprises appear in Table 1.4 of this report.
Problems with Prescription Drug Benefits and Coverage
*= significantly different from
the national average
Mean
National Distribution

State Distribution:

N=26,654

12.2

N=858

12.8

Note: N/A means too few disenrollees answered the questions to permit reporting. Percentages may not add to 100 due to
rounding. The mean score is calculated as the average percentage of reasons endorsed in this composite (0 to 100 scale).
Contract means appearing in bold have adequate reliability (0.70 or above). Contract means not appearing in bold have low
reliability (0. 60-0.70). Contract means appearing in brackets have very low reliability (below 0.60) but are statistically
significantly different from the national mean. For information on how we tested for statistical significance, assessed reliability,
and adjusted for case-mix, see Section IV of this report, pp. 18-22.

IV.

BACKGROUND AND METHODOLOGY

Since 2000, CMS has been collecting information on beneficiaries’ experiences with health care
for Medicare managed care and traditional fee-for-service (FFS) Medicare through the Medicare
Consumer Assessment of Healthcare Plans and Systems (MCAHPS) survey. In 2007, a new
section was added to the survey to assess prescription drug plans under the new Medicare Part
D benefit, including both MA-PDs and PDPs.
In 2012, CMS launched the Medicare Advantage and Prescription Drug Plan Disenrollment
Reasons Survey to capture the reasons Medicare beneficiaries voluntarily disenroll from their
Part C and Part D contracts. Voluntary disenrollment occurs when a beneficiary either drops
coverage entirely or switches to another contract. The survey excludes beneficiaries who
involuntarily disenrolled from contracts for eligibility reasons, moved out of their contract’s
service area, switched benefits packages within your contract, were involuntarily re-assigned or
passively enrolled in a Medicare-Medicaid Plan, or died.
The survey was tested extensively in a full national implementation conducted between
November 2010 and July 2011. Based on experience with this pilot implementation and
subsequent fielding of the survey, refinements in sampling, question wording, and
12

administration mode were made for the surveys that were fielded between January and
December 2017.

Methodology
The Survey Instrument
The Medicare Advantage and Prescription Drug Plan Disenrollment Reasons Survey
includes three versions, directed respectively at disenrollees in three different types
of plans:
•
•
•

Medicare Advantage-only (MA-Only) plans
Medicare Advantage Health and Drug (MA-PD) plans
Medicare Prescription Drug Plans (PDPs)

The MA-PD Survey contains 63 questions, organized into the following sections:
• Your Former Health Plan (2 questions)
• Getting Information or Help from Your Former Health Plan (6 questions)
• Getting Health Care and the Prescription Medicines You Needed from Your Former
Health Plan (9 questions)
• Reasons You Left Your Former Health Plan (27 questions)
• Other Reasons for Leaving Your Former Health Plan (4 questions)
• About You (15 questions)
For scoring and report ing purpos es, we combined questions on relat ed issues about reasons
for disenrollment into the following five composit e (summary ) measures.

• Problems with Coverage of Doctors and Hospitals
• Financial Reasons for Disenrollment
• Problems Getting Needed Care, Cover age, and Cost Information
• Problems Getting Information and Help from the Plan
• Problems with Prescription Drug Benefits and Coverage
Table 1.5 displays the survey questions that go into each of these composites. Table 1.6
contains individual survey questions that are not part of composites but that may be
helpful for informing quality improvement efforts.

13

Table 1.5. Medicare Advantage Health and Drug Plan Disenrollment Reasons Survey
Composite Measures
Composite Measure
Problems with Coverage of Doctors
and Hospitals
Financial Reasons for Disenrollment

Problems Getting Needed Care,
Coverage, and Cost Information

Problems Getting Information and
Help from the Plan

Problems with Prescription Drug
Benefits and Coverage

Survey Questions Included in the Composite
Q35: Did you leave your former plan because the doctors or other health care
providers you wanted to see did not belong to the plan?
Q36: Did you leave your former plan because clinics or hospitals you wanted
to go to for care were not covered by the plan?
Q20: Did you leave your former plan because the dollar amount you had to
pay for each time you filled or refilled a prescription went up?
Q22: Did you leave your former plan because the monthly fee went up?
Q24: Did you leave your former plan because you found a health plan that
costs less?
Q25: Did you leave your former plan because a change in your personal
finances meant you could no longer afford the plan?
Q32: Did you leave your former plan because you were frustrated by the
plan’s approval process for care, tests, or treatment?
Q33: Did you leave your former plan because you had problems getting the
care, tests, or treatment you needed?
Q34: Did you leave your former plan because you had problems getting the
plan to pay a claim?
Q37: Did you leave your former plan because it was hard to get information
from the plan – like which health care services were covered or how much a
specific test or treatment would cost?
Q30: Did you leave your former plan because you did not know whom to
contact when you had a problem filling or refilling a prescription?
Q31: Did you leave your former plan because it was hard to get information
from the plan – like which prescription medicines were covered or how much
a specific medicine would cost?
Q38: Did you leave your former plan because you were unhappy with how
the plan handled a question or complaint?
Q39: Did you leave your former plan because you could not get the
information or help you needed from the plan?
Q40: Did you leave your former plan because their customer service staff did
not treat you with courtesy and respect?
Q23: Did you leave your former plan because they changed the list of
prescription medicines they cover?
Q26: Did you leave your former plan because the plan refused to pay for a
medicine your doctor prescribed?
Q27: Did you leave your former plan because you had problems getting the
medicines your doctor prescribed?
Q28: Did you leave your former plan because it was difficult to get brand
name medicines?
Q29: Did you leave your former plan because you were frustrated by the
plan’s approval process for medicines your doctor prescribed?

14

Table 1.6. Individual Survey Questions Not Included in a Composite Measure*
Individual Survey Questions Not
Included in a Composite
Co-payment for Doctor Visit Went Up
Low Medicare Star Rating
Found Plan with a Higher Medicare
Star Rating
Family Member or Friend
Recommended Another Plan
Saw Commercial or Advertisement for
Another Plan
Another Plan Better Met Prescription
Needs
Another Plan Offered Better Benefits
or Coverage of Health Services

Q21: Did you leave your former plan because the dollar amount you had to
pay each time you visited a doctor went up?
Q42: Did you leave your former plan because it got a low Medicare Star
Rating?
Q43: Did you leave your former plan because you found another plan with a
higher Medicare Star Rating?
Q45: Did you leave your former plan because a family member or friend told
you about a better plan?
Q46: Did you leave your former plan because you saw a commercial or
advertisement for a health plan you thought you would like better?
Q47: Did you leave your former plan because you found another plan that
better met your prescription needs?
Q48: Did you leave your former plan because another plan offered better
benefits or coverage for some types of care, treatment or services (for
example, dental or vision care)?

* These questions were analyzed individually because responses to the questions were not strongly related to responses to
questions included in the composites presented in Table 1.5.

Calculation of Composite Means
Your contract’s mean on a composite measure is calculated as the average percentage of
reasons endorsed in the composite multiplied by 100. To understand this calculation, consider a
composite measure comprised of four survey questions (i.e., reasons for disenrollment).
Suppose that 150 disenrollees from your contract answered these questions, and that 60 of
those disenrollees endorsed none of the four reasons in the composite (0% of all reasons in the
composite), 40 disenrollees endorsed 1 of the 4 reasons (25% of all reasons in the composite),
25 disenrollees endorsed 2 of the 4 reasons (50% of all reasons in the composite), 15
disenrollees endorsed 3 of the 4 reason (75% of all reasons in the composite), and 10
disenrollees endorsed all four reasons (100% of all reasons in the composite). In that case, the
average percentage of reasons in the composite that were endorsed by disenrollees from your
contract would be 29.2% or [(60*0) + (40*0.25) + (25*0.50) + (15*0.75) + (10*1.00)]/150.
Multiplying this average percentage by 100 would yield your contract’s mean score on the
composite: 29.2.

Reporting of Composite Means on the Medicare Plan Finder
The Medicare Plan Finder on the Medicare.gov website (https://www.medicare.gov/find-a-plan)
displays your contract’s mean scores on the composite (summary) measures of disenrollment in
addition to your contract’s voluntary disenrollment rate. These scores are publicly reported so
that consumers making plan choices can see the reasons that beneficiaries gave for disenrolling
from a contract in 2017.
Unlike the results shown in this report, which presents scores to 1 decimal digit, the Medicare
15

Plan Finder presents scores rounded to the nearest integer (whole number). Table 1.7 shows
your contract’s composite scores to 1 decimal digit (as they are presented in this report) and to
the nearest integer (as they are presented on the Medicare Plan Finder). Table 1.7 also shows
the labels used for the composite measures on the Medicare Plan Finder which are different
(i.e., more consumer friendly) than the labels used to describe composite measures in this
report.
In rounding decimals to integers, we followed these standard rules: If the number beyond the
decimal is less than 5, it is rounded down to the next whole number; if the number beyond the
decimal is 5 or more, it is rounded up to the next whole number. This can occasionally lead to
apparent discrepancies in the table below even though the rounding rules have been properly
applied. For example, a score of 19.46 would get rounded to 19.5 to produce a score for this
report, but it would get rounded down to 19 to produce a score for the Medicare Plan Finder.
Table 1.7. Composite Measure Labels and Mean Scores as
Presented in This Report vs. the Medicare Plan Finder
This Report

Composite Measure Label
Problems with Coverage of Doctors
and Hospitals
Financial Reasons for Disenrollment

Your Contract
(
) Score
to 1 Decimal
Digit

Problems Getting Needed Care,
Coverage, and Cost Information
Problems Getting Information and
Help from the Plan*
Problems with Prescription Drug
Benefits and Coverage

Medicare Plan Finder

Composite Measure Label
Doctors or Hospitals are not
Covered by the Plan
Problems with the Cost of the
Plan
Problems Getting the Plan to
Provide and Pay for Needed
Care
Problems Getting Information
and Help from the Plan
Problems with the Plan’s
Prescription Drug Coverage

Your Contract
(
) Score
Rounded to
Nearest Integer

* In previous reports, this measure was labeled "Problems Getting Information about Prescription Drugs"

Sample Selection and Eligibility Criteria
The survey was intended to repres ent the populat ion of beneficiaries who disenrolled
volunt arily from Part C or Part D contracts during the period January 2017 through December
2017. Samples were drawn to repres ent that population with targets of 75 completed
respons es cases from each MA contract's disenrollment population for that period and 150
from each PDP; target sample sizes were calculated based on historical respons e rates from
each contract and also took into account differenc es in respons e rate associated with dual
eligibility and month of disenrollment. Becaus e benefic iaries who disenroll at different times of
the year tend to do so for different reasons and have somewhat different characterist ics, a

16

further goal of the sample design was to be repres ent ativ e of the distribution of each contract ' s
disenrollment across months of the year. Sampling was done month-by -mont h over the course
of the year rather than retrospectively, so the number of cases to be sampled each month
had to be calculated before disenrollment counts from later months were known. Monthly
allocations were projected based on historical patterns of the distribution of disenrollment over
months and adjusted each month as new disenrollment data were received.
For some contract s, the target number of responses is not reached over the course of the year.
Contracts projected to have more than 40 disenrollees over the course of the year but fewer than
the target number of disenrollees ("small" contracts) are sampled at a rate of 100%. Contract s
that were projected to fall below 40 cases per year were excluded from sampling for contractlevel reporting. For these excluded contract s, a floor sampling rate was established to achieve
national (but not contract-level) representativeness.

For national representativeness, sample sizes were increased in the largest (“big”) contracts,
using sampling rates of 0.30% in PDP contracts and 1.50% in MA contracts when these rates
yielded larger samples than the targets described above. The terms “small” and “big” here are
defined with respect to disenrollment, not enrollment.
In MA contracts with some but not all beneficiaries enrolled in the prescription drug (PD)
benefit, samples were drawn from both the MA-PD and MA-Only portions of the contract, and
each group of sampled disenrollees was mailed the appropriate questionnaire.

Survey Implementation
The 2017 survey of disenrollees was conducted with beneficiaries who disenrolled between
January and December 2017. It asked about beneficiaries’ experiences with their plan and
reasons for disenrollment. Data were collected on an ongoing basis and as close as possible to a
beneficiary’s actual date of disenrollment to help with respondent recall. The majority of
voluntary disenrollment occurs in December of each year. Surveys mailed in March 2017
covered disenrollments that happened in the January 2017 time period (approximately a sixweek lag). Surveys mailed in April 2017 covered disenrollments that happened in February 2017,
and so forth. The data collection protocol included mailing of pre-notification letters and up to
two mailings of paper surveys. Residents of Puerto Rico received a Spanish-language version of
the survey. All other disenrollees received an English-language version of the survey but could
request a Spanish language version by calling a 1-800 number maintained by the survey vendor.

Sample Disposition
The sample disposition and response rates for the 2017 Medicare Advantage and Prescription
Drug Plan Disenrollment Reasons Survey are presented in Table 1.8. Of the 122,746 MA
disenrollees in the original sample, 4,911 (4.0%) were classified as ineligible because they were
17

institutionalized, deceased, mentally or physically incapable of responding, or had a language
barrier that prevented them from completing the survey. Eligible sample members who refused
to take the survey or could not be contacted were considered non-respondents (66.3% of
sample members). The adjusted response rate, after accounting for ineligible sample members,
is 30.9% percent (36,426 partial or completed surveys divided by 122,746 disenrollees in the
original sample minus 4,911 disenrollees deemed ineligible).
Table 1.8. 2017 Medicare Advantage and Prescription Drug Plan Disenrollment Reasons
Survey, MA Sample Disposition
Disposition
Partial or completed survey
Completed mail survey
Screened out or did not answer any
substantive items
Ineligible
Deceased
Mentally or physically unable to respond
or institutionalized
Did not speak English or Spanish
Otherwise excluded from survey*
Non-respondent s
Total sample

MA Sample Member Count
36,426
27,899
8,527

Percentage of Sample
29.7%
22.7%
6.9%

4,911
67
28

4.0%
0.1%
0.0%

16
4,800
81,409
122,746

0.0%
3.9%
66.3%
100.0%

* These were mainly surveys returned as undeliverable

Weighting and Case-Mix Adjustment
The survey data are adjusted using regression to remove predictable effects of respondent
characteristics on survey scores and therefore make comparisons among contracts fairer. The
scores represent case-mix adjusted estimates, which approximate the scores that would be
obtained if the respondents from each contract had an identical distribution of characteristics.
Adjustments to the Distribution of Disenrollment by Month and Dual Eligibility to
Obtain Contract- Representative Scores
In 2017, we adjusted contract-representative estimates for differences in the rate of valid
responses per disenrollee by dual status and month of disenrollment. Dual eligibility is an
important variable to consider in any analysis of Medicare experience-of-care data—particularly
where beneficiary financial factors are under consideration—because the out-of-pocket cost
structure for duals is so different from that for nonduals. Duals differ from nonduals in other
relevant ways as well, including income, education, and response rates. Furthermore, because
rates of dual eligibility differ tremendously across contracts (from 0 to 100%), the impact of any
such differences at the individual level is greatly magnified. The month of disenrollment is
important because it plays a major role in sample design and because disenrollments in
December fall under an open enrollment period and consequently constitute around 60% of the
18

year’s disenrollments, with potentially different predominant reasons than in other months (see
pp. 23-25 of this report). For this purpose, we aimed to be representative of the distribution of
disenrollments between December (which accounts for the majority of annual voluntary
disenrollment) and the rest of the year. Furthermore, the fraction of disenrollees who are dually
eligible is much higher in January to November than in December because disenrollees who are
dually eligible are much more likely to switch enrollments outside the open enrollment period
than disenrollees who are non-dually eligible.
This adjustment was performed using a combination of weighting and regression adjustment.
The weighting adjustment is less model-dependent than the regression adjustment, reducing the
potential bias if differences in reasons related to dual eligibility or month of disenrollment differ
by contract. A disadvantage of weighting is that it tends to increase the variance of estimates,
giving them a variance equivalent to that of an unweighted estimate based on a smaller sample
size (i.e., the effective sample size). Consequently, the weighting adjustment was applied first
but was constrained not to reduce the effective sample size of estimates based on the entire
sample to less than 75 in MA contracts or 150 in PDP contracts. For each contract, the sequence
of weighting schemes considered was (1) equal weights [chosen for 149 contracts], (2) halfadjustment to matching dual eligible rate [129 contracts], (3) matching dual status distribution
to population total [42 counts], and (4) adjusting to represent the population distribution across
the 4 dual-status by month cells [63 contracts]. In the equal-weighting scheme, each response
within a contract received a weight equal to the number of disenrollees for the contract-year
divided by the number of survey respondents. “Half-adjustment” means that the weights were
the mean of the equal weights and the dual-status weights. An additional constraint was that
weightings were not adopted if they moved the percent dually eligible in the contract further
away from the overall mean rate (about 35%), because contracts with outlying values of
percent-dually eligible were the hardest to match for the comparative analyses that are the
primary objective of this phase. Thus, a small sacrifice of reduced effective sample size is traded
off for the benefit of bringing the (weighted) distribution of respondents closer to the
distribution of the entire population, with the remaining discrepancies adjusted for by
regression adjustment as part of the case-mix adjustment procedure described below.
Adjustment to Address Discrepancies between Each Contract and the Average Contract in
Terms of the Characteristics of Enrolled Beneficiaries (Case-Mix Adjustment)
An additional adjustment was made to address the effects of discrepancies between the
distributions of characteristics of enrolled beneficiaries in each contract and the average
distribution. Analyses of CAHPS data have shown that beneficiaries with certain characteristics
tend to report more favorable or less favorable experiences even when they are members of the
same contract and have therefore been exposed to the same level of contract quality. Notably,
older plan members, healthier plan members, less educated plan members, and those with
lower socioeconomic status (SES) tend to assess their experiences more favorably than younger,
19

sicker, more educated members and those with higher SES. Similar effects were observed in
analysis of reasons for disenrollment. Contracts do not all have the same distribution ("case
mix") of enrollees with these characteristics, so these response tendencies can bias comparisons
among contracts. If a contract has a large number of enrollees whose characteristics make them
a "tough audience," the contract may receive a lower score than it would receive if it delivered
the same care to enrollees with an average distribution of characteristics.
We perform "case-mix adjustment" using linear regression to correct for these effects and
estimate the scores that would be obtained by each contract if every contract had the same
distribution of enrollee characteristics, equivalent to the average across all contracts. This
analysis was performed after the weighting adjustment that moved the distribution by month
(January-November vs December) closer to the contract’s actual proportions by month, subject
to an effective sample size constraint (see above). The regression adjustments applied these
weights to all analyses. The overall national mean (weighted by number of disenrollees) is the
same before and after adjustment, so scores for some contracts (those with beneficiaries who
tend to provide more favorable assessments) were adjusted downwards relative to their
unadjusted weighted means, and others were adjusted upwards.
The following variables are used in case mix adjustment:
•

Age: A self-reported six-category survey variable ranging from 18 to 85 plus;

•

Education: A self-reported six-category survey variable ranging from less than eighth
grade to more than college

•

Self-reported general health status: Five-category variable (excellent, very good, good,
fair, poor)

•

Self-reported mental health status: Five-category variable (excellent, very good, good,
fair, poor)

•

Proxy assistance: Included as two indicators—one for receiving any proxy assistance and
one for a proxy answering questions for the respondent

•

Low Income Supplement (LIS) eligibility and Medicaid dual eligibility: Three-category
variable (dual eligible, non-dual/LIS, non-dual/non-LIS), The LIS and dual eligible variables
came from the 2017 CMS enrollment files.

Assessing Reliability of Scores
For each composite measure, criteria based on inter-unit reliability (IUR) were applied to classify
each contract’s data as acceptable, low or very low reliability. Inter-unit reliability is defined by
IUR=s2/(SE2+s2), where s2 = between-contract model variance, and SE = standard error of
contract mean. IUR may be interpreted as the fraction of variation in contract mean scores
(among those with about the same IUR) that is attributable to actual differences among
contracts (“signal”) rather than sampling variability (“noise”). Thus, IUR close to 1 indicates that
20

sampling variability is negligible, while IUR close to 0 means that we are unable to detect any
variation among contracts and differences in the data are only random error. Contracts for
which IUR<0.70 are considered to have low reliability. Contracts for which IUR < 0.60 are
considered to have very low reliability.
Reliability of the estimates also is affected by a number of other factors, including the fraction of
the contract’s respondents who are eligible to answer a question based on their experiences,
the variability of responses within the contract, and the amount by which contracts differ from
each other nationally on that measure. Reliability summarizes the influence of these factors on
the precision with which a contract’s score can be compared to national distributions.
Within a given measure, low-reliability scores typically are those with fewer respondents, or
possibly with more variability in their responses. Across measures, more low-reliability scores
will be reported for measures with fewer responses (more respondents for whom the measure
does not apply), less variation in scores across contracts, and more variability in scores within
each contract.

Significance Testing
For composite measures of reasons for disenrollment, where scores are the mean percentage of
reasons in the composite that were endorsed multiplied by 100, two-tailed t-tests were used to
assess whether the case-mix adjusted mean for each contract differed significantly from the
overall mean for all contracts in the nation. When contract scores are significantly different from
the national mean at the p<0.05 level, this is noted in the last column of Table 1.3 under
"Different from National Average?" A "No" entry in this column means that the contract's score
was not significantly different from the national average, "Higher" means that it differs
significantly from the national mean and is higher, and "Lower" means that it differs significantly
from the national mean and is lower.
Contract means that have adequate reliability (0.70 or higher in a 0-1 range) are shown in bold.
Contract means that have low reliability (between 0.60 and 0.70) but meet the minimum sample
size are shown without bolding. Contract mean scores with very low reliability (below 0.60) are
shown only when the contract mean is statistically significantly lower or higher than the national
average (p<0.05). In this case, there is strong evidence that the contract is below or above
average but the degree to which it is below or above average is imprecisely measured. For that
reason, the contract mean is reported but it is shown in brackets. Contract means that have very
low reliability (below 0.60) and that are not statistically significantly different from the national
average are suppressed because there is not strong evidence that the contract is distinguishable
from the national average.
Table 1.4 reports results for individual questions, again using the last column on the
right to indicate whether a score differed significantly from the national mean.
21

In Section III of the report (Detailed Results for Composite Measures), results are displayed
graphically for composite measures. When a contract’s score is significantly different from the
national average, it is noted by an asterisk next to the score and with text at the top of the
figure.
To compare the contract's disenrollment rate with the national average, we used a chi-square
test. To compare rates of endorsement of individual questions underlying the composites with
the national average, we used a t-test on case-mix-adjusted results. For all tests performed,
differences that are significant at the p<0.05 level are noted in the table or figure as described
above.

Comparison of Reasons for Disenrollment: January-November versus December
Disenrollees
Analyses of results from the Medicare Advantage and Prescription Drug Plan Disenrollment
Reasons Survey have shown that the reasons why beneficiaries report disenrolling from their
health or prescription drug plans differ between those who disenroll in the January-November
months as compared to those who disenroll in December. Individuals who disenroll during the
Open Enrollment Period, which runs from October 15 to December 7, show up as December
disenrollees. Individuals who disenroll outside of the Open Enrollment Period, and thus show up
as January-November disenrollees, are beneficiaries who qualify both for Medicare and
Medicaid (i.e., dual eligibles) or who qualify for a Special Election Period (SEP) (i.e., non-duals
who qualify for the Low Income Subsidy). Table 1.9 shows national average scores on the
composite measures of disenrollment as well as on the questions that make up those composite
measures broken down between January-November disenrollees and December disenrollees.
Table 1.10 shows a similar breakdown of national average scores on the individual survey
questions that are not included in a composite. Contract-specific scores for January-November
and December disenrollees are not reported because of insufficient sample size. Nevertheless, it
may be useful to keep the differences shown below in mind when interpreting your own
contract’s scores.

22

Table 1.9. National Average Scores on Composite Measures and Their Constituent
Questions: January-November Disenrollees vs. December Disenrollees

Problems with Coverage of Doctors and Hospitals
Preferred provider not in plan
Preferred clinic or hospital not covered by plan
Financial Reasons for Disenrollment
Found a plan that costs less
Prescription co-payment went up
Monthly premium went up
Could no longer afford plan
Problems Getting Needed Care, Coverage, and Cost Information
Problems getting needed care, tests, or treatment
Frustration with approval process for care, tests, or treatment
Hard to get information about coverage and cost of health services
Problems getting claims paid
Problems Getting Information and Help from the Plan
Could not get information or help needed from the plan
Unhappy with how the plan handled a question or complaint
Hard to get information about coverage and cost of prescription drugs
Customer service not courteous or respectful
Did not know whom to contact about filling a prescription
Problems with Prescription Drug Benefits and Coverage
Frustrating approval process for off-formulary medications
Problems getting prescribed medication
Plan refused to pay for a prescribed medication
Change in drug formulary
Difficult to get brand name medications

National Average
January-Nove mbe r
Disenrollees

December
Disenrollees

34.4
26.8

29.6
20.6

32.5
18.5
14.5
19.2

46.1
21.9
22.9
15.7

28.8
27.0
17.3
14.6

19.1
20.4
12.2
10.5

25.2
23.3
12.5
10.2
8.9

16.7
15.5
9.3
5.7
4.9

16.3
15.8
15.5
12.4
12.9

11.4
10.5
10.3
10.5
9.4

Note: Individuals who disenroll during the Open Enrollment Period, which runs from October 15 to December 7, show up as
December disenrollees. Individuals who disenroll outside of the Open Enrollment Period, and thus show up as January-November
disenrollees, are predominately beneficiaries who qualify both for Medicare and Medicaid (i.e., dual eligibles) or who qualify for a
Special Election Period (i.e., non-duals who qualify for Low Income Subsidy)

Table 1.10. National Average Scores on Single Questions (not in a composite): JanuaryNovember Disenrollees vs. December Disenrollees

Other Reasons for Disenrollment
Another plan offered better benefits or coverage of health services
Another plan better met prescription needs
Family member or friend recommended another plan
Saw commercial or advertisement for another plan that looked better
Co-payment for doctor visit went up
Found a plan with a higher Medicare Star rating
Low Medicare Star rating

National Average
January-Nove mbe r
Disenrollees

December
Disenrollees

49.7
37.0
29.4
23.6
20.8
18.3
7.6

48.7
33.0
32.8
23.9
18.5
13.9
3.3

Note: Individuals who disenroll during the Open Enrollment Period, which runs from October 15 to December 7, show up as December
disenrollees. Individuals who disenroll outside of the Open Enrollment Period, and thus show up as January-November disenrollees,
are beneficiaries who qualify both for Medicare and Medicaid (i.e., dual eligibles) or who qualify for a Special Election Period.

23

State or Regional Comparisons
In addition to comparing your contract's results with a national benchmark, it may be useful to
compare the results with a state or regional benchmark. We have provided such a benchmark
for your contract. For most contracts, the benchmark is the state with the largest number of
enrollees in that contract in 2017. However, we used broader regional benchmarks (census
divisions) instead of states when any of the following occurred:
•

The state had no more than one contract with 10 or more respondents represented in
the disenrollment survey

•

A single contract with large market share accounted for 75% or more of the
disenrollments in the state

Under these conditions, state-level benchmarks would not be meaningful or useful.
Because sample sizes for state and regional benchmarks are much smaller than for national
benchmarks, we do not provide statistical tests for these comparisons.

Contact Information
If you have questions about the survey or this report, please send them to
DisenrollSurvey@cms.hhs.gov.

24