Supporting statement A_CMS-10316_Disenrollment Reasons Survey_CLEAN V2

Purpose of the survey: The Medicare Prescription Drug Plan (PDP) and Medicare Advantage (MA) Plan Disenrollment Reasons Survey (Disenrollment Survey) focuses on beneficiaries who voluntarily disenroll from their plan. Beneficiaries can disenroll from plans during the Annual Election Period (AEP), the Medicare Advantage Disenrollment Period

(MADP), and Special Election Periods (SEPs). The Centers for Medicare & Medicaid Services (CMS) developed the Disenrollment Survey to capture the reasons for disenrollment at a time that is as close as possible to the actual date of disenrollment. Through this survey, CMS seeks to: (1) obtain information about beneficiaries’ expectations relative to provided benefits and services (for both MA and PDPs) and (2) determine the reasons that prompt beneficiaries to voluntarily disenroll. It is important to include such information from disenrollees as CMS assesses plan performance, because plan disenrollment can be a broad indicator of beneficiary dissatisfaction with some aspect of plan services, such as access to care, customer service, cost, benefits provided, or quality of care. Information obtained from the Disenrollment Survey also supports the quality improvement efforts of individual plans and provides data to assist consumer choice through use of the Medicare Plan Finder website. (Note, when we refer to a “plan” we focus on disenrollment from what CMS calls a contract, or H, R, or S number, not changes at the plan benefit package level.)

Each year, CMS uses the overall rate of disenrollment from MA and PDP plans as a performance measure in the annual Star Rating program for Part C and Part D contracts. In 2017, among plans with at least 1,000 enrolled members, 3.8 million beneficiaries voluntarily disenrolled from a plan, with disenrollment rates varying widely across contracts. MA disenrollment rates ranged from <1% to 59% and PDP disenrollment rates from 2% to 35%. The Disenrollment Survey extends the overall disenrollment rate to investigate disenrollment reasons nationally, by market/regions, by population subgroups (e.g., beneficiaries who are dually eligible for Medicare and Medicaid vs. non-duals), and by specific plans (i.e., contracts).

The survey is conducted annually and is ongoing. . CMS uses the information obtained

from the Disenrollment Survey for several purposes. The survey results are an important plan monitoring tool for CMS to ensure that Medicare beneficiaries are receiving high quality services from contracted providers. CMS uses information from the survey to track changes in the reasons Medicare beneficiaries cite for disenrolling to monitor improvements/declines over time nationally and at the plan level. CMS also uses the disenrollment survey results to support the quality improvement efforts of individual plans, by providing plans with a detailed, annual report showing the reasons disenrollees cited for voluntarily leaving the plan and comparing the plan’s scores to regional and national benchmarks. Additionally, CMS uses the plan-specific results of the survey to provide Medicare beneficiaries with information (i.e., reasons cited for disenrolling from a plan and the frequency with which disenrollees cite each of the reasons) to assist beneficiaries with their annual consumer choice of plans. The disenrollment survey results are shared with beneficiaries on the Medicare Plan Finder website

(https://www.medicare.gov/find-a-plan) where they can compare the performance of plans.

CMS’ survey contractor pulls a monthly sample from CMS’ monthly disenrollment file (which contains the universe of voluntary disenrollments for that month) over a 12-month period for each contract, with the goal of reaching a target sample size per plan of approximately 233 (MA plans) and 465 (PDP plans). CMS does not survey the census of all disenrollees, rather only a sample of disenrollees. The size of the sample was determined by the number needed to generate reliable estimates at the plan-level based on survey response rates, screen-in rates, and variation in responses to survey items. CMS draws a larger sample size for PDP plans due to the fact that there is less variation between PDP plans in the reasons cited for disenrolling, as compared to MA plans. CMS pulls a random sample of disenrollees from each plan each month.

The large sample drawn each year is necessitated by the dual purposes of the survey: to generate national estimates of reasons for disenrollment and to produce reliable plan-level estimates for reporting to plans and Medicare beneficiaries plan-level estimates of reasons for disenrollment. The survey results are intended to represent the population of beneficiaries who disenrolled voluntarily from Part C (MA-Only or MA-PD) or Part D (PDP) contracts during an annual period (i.e., January through December each year). To represent that population, we target an annual sample of 233 cases from each MA contract’s annual voluntary disenrollment and 465 from each PDP. The 233/465 sample sizes were based on analyses the survey contractor performed to determine the number of cases required to achieve reliable estimates (i.e., reliability of 0.70 or greater^¹) of reasons for disenrollment. Beneficiaries who disenroll at different times of the year may tend to do so for different reasons and have somewhat different characteristics (e.g., a higher fraction of dually eligible beneficiaries disenroll in non-December months as they are eligible to disenroll during the Special Election Period); as such, a further goal of the sample design was to represent the distribution of each contract’s disenrollment across months of the year. Sampling is done month-by-month over the course of the year rather than retrospectively once all disenrollment for a contract is known for the year. In each calendar year, we estimate approximately 143,000 sampled cases (plus or minus). Across the

143,000 sampled cases, roughly 114,000 are allocated to disenrollees from MA-Only and MA-

PD contracts (approximately 568 plans), and 29,000 to disenrollees from PDP plans (approximately 62 plans). The total allocated annually varies based on several factors including: 1) the total number of MA and PDP plans, as some plans terminate and new plans enter the market; and 2) fluctuations in screen-out rates and response rates.

Historical Context for the Survey: Voluntary disenrollment rates from managed care plans are often viewed as a good “summary” indicator of member satisfaction and plan quality. The Balanced Budget Act of 1997 required that the CMS publicly report two years of disenrollment rates on all Medicare + Choice (M+C) organizations. To ensure that disenrollment rates would be meaningful to beneficiaries in health plan choice, to support quality monitoring activities, and to assist in quality improvement initiatives, CMS funded the development and implementation of an annual national survey to identify the reasons why beneficiaries voluntarily leave health plans. From 2000 through 2005, CMS administered the Medicare Consumer Assessment of Health Plans (CAHPS) Disenrollment Reasons Survey for managed care organization and publicly reported information from this survey.

As the Medicare program had changed significantly since the CAHPS Disenrollment

Reasons Survey was administered in 2005 (largely attributable to the 2006 implementation of Medicare’s Prescription Drug, or Part D, benefit), CMS funded an effort in 2009 to develop a revised disenrollment survey to focus on beneficiary reasons for voluntarily leaving PDP and

MA plans; the pilot disenrollment survey work occurred between November 2010 - July 2011 (approved under OMB CONTROL NUMBER: 0938-1113 and focused only on beneficiaries who voluntarily disenrolled from their MA-PD or PDP plan, excluding those who involuntarily disenrolled from contracts because of ineligibility, movement out of the contract’s service area, or death. This initial survey effort served as a large-scale field test of methods (data fielding, sampling, weighting, and composite construction), to understand response rates, identify any issues with the survey tool, and examine the most important reasons for disenrollment. Through this work, several improvements were identified, including refinements to survey wording regarding contract name recognition and efficiencies in the administration of the survey, and refining the sample size required to generate reliable contract level estimates of reasons for disenrollment.

Starting in 2012, CMS moved to implementation of the survey on an annual basis to provide annual feedback to plans and to support annual choice of plans by beneficiaries. CMS expanded the survey to include disenrollees from MA-only contracts and continued to work with the survey contractor to improve screen-in rates and respondent comprehension regarding the contract that they have left.

CMS, working with their contractor, designed and distributed reports to 38 PDP and

372 MA-PD contract reports based on calendar year 2017 disenrollment reasons survey data. Since 2012, the survey has been implemented nationwide on an annual basis to generate data to provide CMS with data for contract monitoring, to produce individual plan reports that are used to inform plans’ quality improvement efforts, and to produce information for Medicare beneficiaries to use when selecting plans.

To note, CMS has been collecting information since 2000 on beneficiaries’ experiences with health care for Medicare managed care and traditional fee-for-service (FFS) Medicare among enrollees in plans through the Medicare Consumer Assessment of Healthcare Plans and Systems (MCAHPS) survey. Starting in 2007, the MCAHPS survey added a new section to assess prescription drug plans under the new Medicare Part D benefit and developed a new MCAHPS survey instrument for beneficiaries enrolled in PDPs. Although CMS was collecting the experiences of enrolled members, outside of consumer complaints (i.e., the Medicare Beneficiary Ombudsman and grievance and appeals process), very little was known about the reasons why beneficiaries disenroll from MA and PDP plans, information that could be used to drive improvements in care and services to Medicare beneficiaries. The Disenrollment Survey was designed to fill that information gap.

Survey Content and Composite Measures: The MA and PDP Disenrollment Reasons Survey includes three versions, directed respectively at disenrollees in three different types of plans:

Medicare Advantage-only (MA-Only) plans (45 questions)
Medicare Advantage Health and Drug (MA-PD) plans (63 questions) • Medicare Prescription Drug Plans (PDPs) (54 questions)

As an example, the MA-PD Survey is 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)

CMS combines questions on related reasons for disenrollment into composite (summary) measures, which are reported to plans along with individual survey items. Table 1 describes the composites measures constructed and reported by survey type.

Table 1: Composite Measures by Survey Type

Composite Measures	MA-Only Survey	MA-PD Survey	PDP Survey
Financial Reasons for Disenrollment	Yes	Yes	Yes
Problems Getting Needed Care, Coverage, and Cost Information	Yes	Yes	No
Problems with Coverage of Doctors and Hospitals	Yes	Yes	No
Problems Getting Information and Help from the Plan	No	Yes	Yes
Problems with Prescription Drug Benefits and Coverage	No	Yes	Yes

Table 2 displays how individual survey questions map into the composite measures.

Table 2: Composite Measures and Individual Items that Map to Composites

Composite Measure		Survey Questions Included in the Composite
Financial Reasons for Disenrollment	• • •	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? (Q20) Did you leave your former plan because the monthly fee went up? (Q22) Did you leave your former plan because you found a health plan that costs less? (Q24)
	•	Did you leave your former plan because a change in your personal finances meant you could no longer afford the plan? (Q 25)
Problems with Coverage of Doctors and Hospitals	• •	Did you leave your former plan because the doctors or other health care providers you wanted to see did not belong to the plan? (Q35) Did you leave your former plan because clinics or hospitals you wanted to go to for care were not covered by the plan? (Q36)
Problems Getting Needed Care, Coverage, and Cost Information	• • •	Did you leave your former plan because you were frustrated by the plan’s approval process for care, tests, or treatment? (Q32) Did you leave your former plan because you had problems getting the care, tests, or treatment you needed? (Q33) Did you leave your former plan because you had problems getting the plan to pay a claim? (Q34)
	•	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? (Q37)
Problems Getting Information and Help from the Plan	• • •	Did you leave your former plan because you did not know whom to contact when you had a problem filling or refilling a prescription? (Q30) 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? (Q31) Did you leave your former plan because you were unhappy with how the plan handled a question or complaint? (Q38)
	•	Did you leave your former plan because you could not get the information or help you needed from the plan? (Q39)
	•	Did you leave your former plan because their customer service staff did not treat you with courtesy and respect? (Q40)
Problems with Prescription Drug Benefits and Coverage	Did you leave your former plan because they changed the list of prescription medicines they cover? (Q23) Did you leave your former plan because the plan refused to pay for a medicine your doctor prescribed? (Q26) Did you leave your former plan because you had problems getting the medicines your doctor prescribed? (Q27) Did you leave your former plan because it was difficult to get brand name medicines? (Q28) Did you leave your former plan because you were frustrated by the plan’s approval process for medicines your doctor prescribed? (Q29)

Current OMB/PRA request: CMS received its most recent clearance for the

Medicare Prescription Drug Plan (PDP) and Medicare Advantage (MA) Plan Disenrollment

Reasons Survey on May 8, 2017 (OMB control # 0938-1113). This clearance expires on May

31, 2020. CMS requests a three-year clearance (06/01/2020 through 05/31/2023)from the Office of Management and Budget (OMB) under the Paperwork Reduction Act of 1995 to continue annual fielding the Medicare Prescription Drug Plan (PDP) and Medicare Advantage (MA) Plan Disenrollment Reasons Survey. CMS requests OMB’s approval of the survey included as part of this OMB/PRA request (Attachments VI, VII, and VIII). CMS has reviewed the survey response data to assess the performance of individual items (e.g., response rates, screen-outs, item skipping) which has informed survey revisions. In this OMB/PRA update request, CMS has made minor survey modifications (i.e., dropped two items and added two items English version, dropped 4 items and added two items Spanish version) to the previously approved survey (as shown in our survey item crosswalk document that is included with this OMB/PRA update request). CMS will field the annual survey in the same manner as it has been doing since the last OMB approval May 8, 2017. CMS will continue to pull monthly samples of voluntary disenrollees from each MA and PDP plans to produce annual reports of reasons for disenrollment to use for plan feedback and improvement and beneficiary choice.

B. Justification

B1. Need and Legal Basis

The Balanced Budget Act of 1997 required that the CMS publicly report two years of disenrollment rates on all Medicare + Choice (M+C) organizations. Disenrollment rates are a useful measure of enrollee dissatisfaction with a plan; this information is even more useful when reasons for disenrollment are provided to consumers, insurers, and other stakeholders. Advocacy organizations agree that CMS needs to report disenrollment reasons so that disenrollment rates can be interpreted correctly. (See

https://www.medicarerights.org/pdf/Why-Consumers-Disenroll-from-MA.pdf as an example.) The Disenrollment Survey gives CMS, plans, and beneficiaries important information about the reasons members leave Medicare Advantage and Prescription Drug plans.

Further, the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) provides a requirement to collect and report performance data for Part D prescription drug plans. Specifically, the MMA under Sec. 1860D-4 (Information to Facilitate Enrollment) requires CMS to conduct consumer satisfaction surveys regarding the PDP and MA contracts pursuant to section 1860D-4(d). Plan disenrollment is generally believed to be a broad indicator of beneficiary dissatisfaction with some aspect of plan services, such as access to care, customer service, cost of the plan, services, benefits provided, or quality of care.

The information generated from the disenrollment survey supports CMS’ ongoing efforts to assess plan performance and provide oversight to the functioning of Medicare

Advantage (Part C) and PDP (Part D) plans, which provide health care services to millions of Medicare beneficiaries. Beneficiary satisfaction (as measured in the MCAHPS survey) and dissatisfaction (as measured in the disenrollment survey) with plan performance are both important sources of information for plan monitoring and oversight. The disenrollment survey assesses different aspects of dissatisfaction (i.e., reasons why beneficiaries voluntarily left a plan), which can identify problems with plan operations; performance areas evaluated include access to care, customer service, cost, benefits provided, and quality of care. Understanding how well plans perform on these dimensions of care and service helps CMS understand whether beneficiaries are satisfied with the care they are receiving from contracted plans. When and if plans are found to be performing poorly against an array of performance measures, including beneficiary disenrollment, CMS may take corrective action.

B2. Information Users

This data collection complements the enrollee beneficiary experience data collected through the Medicare Consumer Assessment of Healthcare Providers and Systems (MCAHPS) survey by providing information on the reasons for disenrollment from a Medicare Advantage (with or without prescription drug coverage) or Prescription Drug Plan.

From an operations standpoint, the survey is an important plan oversight monitoring tool for CMS to ensure that Medicare beneficiaries are receiving high quality services from contracted providers.

CMS uses the information obtained from the disenrollment survey to support the quality improvement efforts of individual plans. Annually, CMS provides each plan with a detailed report showing the reasons disenrollees cited for voluntarily leaving the plan and comparing the plan’s scores to regional and national benchmarks. The detailed plan reports are available to

CMS staff. See Attachment IX for a redacted Medicare Advantage Prescription plan report.

The disenrollment survey results are an important source of information used by CMS to monitor plan performance and to identify potential problems with plans (e.g., plans providing incorrect information to beneficiaries or creating access problems). CMS uses the results to monitor the quality of service that Medicare beneficiaries are receiving from contracted providers and to understand beneficiaries’ expectations relative to provided benefits and services (for both MA and PDPs).

CMS uses the information from the Disenrollment Survey to support quality improvement efforts of individual plans by providing them with a detailed, annual report showing the reasons disenrollees cited for voluntarily leaving the plan and comparing the plan’s scores to regional and national benchmarks. The annual plan reports include results on individual survey items as well as composite measures of reasons for disenrollment (i.e., financial reasons, problems with prescription drug benefits and coverage, problems getting information and help from the plan, problems getting needed care, coverage, and cost information, and problems with coverage of doctors and hospitals). Plans also see information showing disenrollment rates among subgroups of their enrolled population (e.g., duals/nonduals, elderly vs. non-elderly disabled). Plans can use the information to guide quality improvement efforts. For example, PDP and MA plans (both MA-Only and MA-PD) can identify problem areas, make changes to the types of medications covered, to beneficiary costs, and to other plan features that impact beneficiaries, and reduce the likelihood of disenrolling. A sample of the plan report is provided in the supplemental documents section of this request.

CMS receives feedback about the utility and use of the disenrollment rates and disenrollment reasons through formal means (i.e., CMS Call Letter) and informally, particularly the dedicated Disenrollment Survey mailbox. Each year after distribution of the report, plans frequently reach out through the Disenrollment Survey mailbox, and request more information about the survey which is a signal that plans are using information from the reports/survey. For example, plans have expressed interest in obtaining greater drill down information on performance of all plans within each plan’s local market (i.e., competitors), along with the state and national benchmarks the report already includes. CMS has also received requests from plans that are consolidating, wanting survey results from the plan they are merging with to assess plan’s weaknesses and strengths; such information can help inform how the newly merged plan will conduct its operations moving forward. We have also received requests from plans where the parent organization (i.e., sponsor) wants to use the data in the Disenrollment Survey plan reports to compare the performance for all their plans in different markets. They’ve asked CMS for more information from CMS to help them conduct analyses related to their business operations. CMS also hears from plans that they are awaiting receipt of the plan reports in late summer, as they consider whether and how to change services and benefits for the upcoming year.

CMS’ survey contractor periodically requests feedback from plans to learn whether the reports are clear and contains information of value to the plans. Plans report that the disenrollment survey reports are very helpful and that information in the reports has helped them identify opportunities for improvement. Below are highlights from plan representative comments:

The national comparison is very useful. I want to make sure that there are not any outliers.
We review the report as a group to identify opportunities for improvement with a particular focus on member retention, problem solving, and quantifying issues.
Use the report to identify opportunities to improve member services and to identify whether there is a product issue.
Look to see where they are high on any measure and then try to link it back to changes made in the previous year.
We identify top 5 reasons why members disenroll
We use the report to do a year by year comparison
We do analysis on a yearly basis on member disenrollment. We look at our own data and the plan report data and use that to project and make changes to retain members.
Information in the report is useful for trending.

Notably, when the Disenrollment Survey was discontinued in 2005, CMS received numerous requests from plans to reinstate the survey so that plans could review findings for quality improvement.

CMS also makes the results publicly available to consumers to help them with their annual review and selection of a Medicare Advantage or Prescription Drug Plan. Consumers can view the overall plan disenrollment rate and the summary composite reasons cited for disenrolling (i.e., financial reasons, problems with prescription drug benefits and coverage, problems getting information and help from the plan, problems getting needed care, coverage, and cost information, and problems with coverage of doctors and hospitals), showing the frequency with which disenrollees cited for each reason. The publicly available results of both the disenrollment rate and disenrollment reasons are designed to assist Medicare beneficiaries’ annual review and plan selection. The data are publicly available on the Medicare Plan Finder website (https://www.medicare.gov/find-a-plan/ and by clicking on the plan-level disenrollment rate).

B3. Use of Information Technology

The survey vendor collects the data via a mail data collection strategy that involves two rounds of mailed surveys. The mailed survey is formatted for data scanning, and data from all returned surveys is scanned into an electronic data file. CMS uses a mailed survey protocol for several reasons: 1) many seniors, especially older beneficiaries, are not routine, facile users of the Internet; 2) CMS does not collect or maintain current email addresses of Medicare beneficiaries that would be required to field a web-based survey; and 3) plans frequently do not have or maintain current email addresses (typically email information is collected and updated by physician offices and not transmitted to plans). Mail surveys are less costly to administer than phone surveys (with the exception of Interactive Voice Recognition (IVR) surveys which typically generate very low response rates).

As background, CMS’ Disenrollment Survey contractor explored innovations that CMS might consider to improve response rates and/or to reduce costs to the federal government associated with fielding the survey. One area explored was the feasibility of conducting a web-based survey. The contractor reviewed published and unpublished literature and held discussions with representatives from two health plans (with substantial Medicare Advantage enrollment) and one provider organizations to understand current experience with or future plans related to using e-mail and electronic means of communicating (including surveying) with Medicare Advantage members.

A recent hospital CAHPS (HCAHPS) experiment conducted by Elliott et al. (2012)^²evaluated the effectiveness of web/mail mode (i.e., a mail invitation to participate by web, with the web-link provided in the invitation letter, with a mail survey available upon request). This study found that the web/mail mode yielded half the response rate of mail mode and produced compositionally different respondents. The Pew Internet and American Life Project (Zickuhr, 2012)^³found that 59 percent of older adults (>65 years of age) used the Internet, though this varied by beneficiary characteristics (e.g., age, education, and income), suggesting that a shift to web-based surveying would have uneven penetration across the Medicare population with low penetration among older and poorer beneficiaries. Shifting to email-based surveys presents challenges among the senior population, in that seniors may not use the Internet because they never learned to, because of physical problems, or because they cannot afford the technology.

Added to this, discussions with health plans and the medical group revealed barriers to email communication, including surveys. Health plans said they were interested in using email and the Internet to communicate with and survey their Medicare enrollees, but currently do not. Both plans contacted rely primarily on snail-mail for general communication and phone and mail for surveys. A key hurdle cited by the plans for moving toward web-based communication is that they do not have complete, up-to-date e-mail addresses for their Medicare members. One plan estimated that about 30 percent of their current MA-Only and

MA-PD enrollees gave an e-mail address when they enrolled, 65 percent of which were valid.

Another problem cited by plans is that when they contact MA members by e-mail, only about 10-50 percent of e-mails are opened depending on the subject line. The medical group reported they collect and regularly update e-mail addresses, although they were unsure for how many Medicare members they had valid e-mail addresses. The medical group does not share email addresses with the plans it contracts with.

B4. Duplication of Efforts

This is the only disenrollment reasons survey sponsored by CMS being fielded currently to recent disenrollees from MA and PDP plans. CMS’s disenrollment survey contractor has talked to plans (<9 plans) at various points between 2013 and 2017 regarding the design of the disenrollment survey plan report; during these discussions, almost no plan reported attempting to field their own disenrollment surveys. CMS also has not received questions from disenrollees or plans about multiple surveys.

B5. Small Businesses

Survey respondents are Medicare beneficiaries who disenroll from Medicare

Prescription Drug Plans (PDPs), and Medicare Advantage plans (both MA-Only plans and MA-PD plans). The survey should not impact small businesses or other small entities.

B6. Less Frequent Collection

The consequence of not collecting data as soon as possible after a beneficiary disenrolls from a health or prescription drug plan is that the beneficiary will be less able to recall their specific reasons for disenrolling from a PDP or MA plan and their experiences under their previous plan, information that is critical for program improvement. PDP and MA plans (both MA-Only and MA-PD) can make changes to the types of medications covered, to beneficiary costs, and to other plan features that impact beneficiaries. It is therefore useful that CMS survey on an ongoing monthly basis, sampling from the most recent set of disenrollees to enhance recall as to the reasons for disenrollment and the plan the beneficiary has disenrolled from.

Further, it is important that plans, beneficiaries and CMS have access to recent information on reasons for disenrollment to guide decision making. In the last OMB approval, as a term of clearance, OMB requested that when four years of data are available, CMS evaluate the within plan temporal variability in quality scores available to consumers and adjust the frequency of the data collection accordingly. Additionally, OMB requested that CMS look at the temporal and geographic variability in the distribution of disenrollment reasons across all plans (analyses will include comparisons at the 10th, 25th, 50th, 75th, 90th percentiles) to assess changes over time and the utility of annual surveying.

In response to this request from OMB, CMS examined the following:

contract-level variability in reasons for disenrollment composite scores over time;
geographic-level variability in the distribution of disenrollment reasons over time; and
the degree to which disenrollment reasons composites predict future disenrollment.

The analyses (details of which can be found in Attachment X) show that annual survey data collection provides timely, contract-specific information to inform contract quality improvement and beneficiary choice of contracts. Because MA and PDP contracts often change from year-to-year in the services and plan benefit package options they provide, it is important that performance data be as current as possible for consumers to make decisions and for contracts to understand how their members rate their performance. We find that performance and disenrollment reason information vary sufficiently year-to-year as to capture information that may not be well represented by using data from even one year before, much less two or three years before. Given the observed changes in the quality scores both at the contract-level and regionally (i.e., changes in percentile categories identified by OMB) across time, there is significant utility to CMS, contracts, and beneficiaries in continuing to collect the disenrollment reasons information on an annual basis. The findings of variability across years between contracts and between states suggest the value of frequent administration and reporting of the disenrollment reasons survey, given that there are often substantial changes for contracts and states in one year. Additionally, there is a demonstrable, predictive link between the proportion of non-financial reasons cited by disenrollees in one year and the subsequent year’s disenrollment rate for a contract; this highlights the utility of providing the most recent disenrollment reasons information to contracts, who are focused on minimizing disenrollment.

B7. Special Circumstances

This collection doesn’t contain any special circumstances.

B8. Federal Register/Outside Consultation

The Agency’s 60-day Federal Register notice was published in the Federal Register (84 FR 55966) on 10/18/2019. In response to public comments (attached), we removed 6 items and added 2 items to the MA-PD survey, removed 5 items from the PDP survey, and removed one item and added 2 items to the MA-Only survey. Proposed changes to the 2020 Disenrollment Survey instruments are detailed in the attached survey crosswalk and briefly described here, as well as in the response to public comments document.

CMS proposes removing the following patient experience items:

“Did you ever try to get information from your former plan about which prescription medicines were covered?” (MA-PD Q5; PDP Q5)
“How often did your former plan give you all the information you needed about which prescription medicines were covered?” (MA-PD Q6; PDP Q6)
“Did you ever try to get information from your former plan about how much you would have to pay for a prescription medicine?” (MA-PD Q7; PDP Q7)
“How often did your former plan give you information about how much you would have to pay for a prescription medicine?” (MA-PD Q8; PDP Q8)
“Did you ever try to get any kind of care, tests, or treatment through your former plan?” (MA-PD Q9; MA-Only Q5)
“Did a doctor ever prescribe a medicine for you that your former plan did not cover?” (MA-PD Q11; PDP Q9)

And, CMS proposes adding the following two items to the MA-PD and MA-Only surveys:

“In the last 6 months, did you make an appointment to see a specialist?”
“In the last 6 months, how often did you get an appointment to see a specialist as soon as you needed?”

CMS proposes these changes to align the patient experience items between the disenrollment survey and the MA and PDP CAHPS survey. The two new items will be used to compare the experiences between enrollees and disenrollees and provide a greater understanding of why members leave a plan. These revisions do not change any requirements and reduce the overall burden estimates.

The Agency’s 30-day Federal Register notice was published in the Federal Register (85 FR 16634 ) on 03/24/2020,.

B9. Payment/Gifts to Respondents

None. This data collection will not include respondent incentive payments or gifts. Plans are provided with the results of the survey in the form of an annual report for their use in making quality improvements. Beneficiaries are provided with the composite measure survey results on Medicare Plan Finder.

B10. Confidentiality

Individuals contacted as part of this data collection will be assured of the confidentiality of their replies under 42 U.S.C. 1306, 20 CFR 401 and 422, 5 U.S.C. 552 (Freedom of Information Act), 5 U.S.C. 552a (Privacy Act of 1974), and OMB Circular A-130.

B11. Sensitive Questions

The survey does not include any questions of a sensitive nature.

B12. Burden Estimate (Hours & Wages)

Wages

To derive average costs, we used data from the U.S. Bureau of Labor Statistics’ May

2018 National Occupational Employment and Wage Estimates for all salary estimates (https://www.bls.gov/oes/current/oes_nat.htm#00-0000). In this regard, the following table presents the mean hourly wage for “All Occupations,” the cost of fringe benefits (calculated at 100 percent of salary), and the adjusted hourly wage.

Occupation Title	Occupation Code	Mean Hourly Wage ($/hr)	Fringe Benefit ($/hr)	Adjusted Hourly Wage ($/hr)
All Occupations	00-0000	$24.98	$24.98	$49.96

We adjusted our employee hourly wage estimates by a factor of 100 percent to account for fringe benefit costs, which is a rough adjustment because fringe benefits and overhead costs vary significantly from employer to employer, and methods of estimating these costs vary widely from study to study. Nonetheless, there is no practical alternative and we believe that doubling the hourly wage to estimate total cost is a reasonably accurate estimation method.

Burden Estimates

Exhibit 1 shows the estimated annualized burden for the respondents' time to participate in this data collection. The Medicare Prescription Drug Plan (PDP) and Medicare Advantage (MA) Plan Disenrollment Reasons Survey will be administered to 142,431 beneficiaries in calendar- year 2019 (113,260 MA disenrollees and 29,171 PDP disenrollees) using three survey versions:

Medicare Advantage with Prescription Drug Plan Coverage or MA-PD version (Attachment VI in Supporting Statement B); (2) Stand Alone Prescription Drug Plan or PDP version (Attachment VII in Supporting Statement B); and (3) Medicare Advantage Only or MA-Only version (Attachment VIII in Supporting Statement B). We anticipate an annual overall response rate of approximately 30-32% given multiple previous years’ experience. In 2017, MA response rates were 29.9% and PDP response rates were 31.2%, for a total overall response rate of 29.9%.The estimated average response time of 0.18 hours or 11 minutes for the PDP version of the survey is based on the length of that survey version, a pace of 4.5 items per minute, standardized survey instructions, and CMS’ experience with surveys of similar length that were fielded with Medicare beneficiaries. Similarly, the estimated average response time of 0.22 hours or 13 minutes for the two MA versions of the survey (MA-PD and MA-Only) is based on the length of the MA-PD survey version, a pace of 4.5 items per minute, and CMS’ experience with surveys of similar length that were fielded with Medicare beneficiaries.

Note: although the MA-Only survey instrument is shorter than the MA-PD survey instrument (46 vs. 59 items), for this burden estimate we are assuming that all MA disenrollees will fill out the MA-PD (longer) version because there are only a very small number of MA-Only plans and we know that MA-Only surveys make up a minority of the total MA sample. As indicated below, the total burden hours are estimated to be 9,354 hours.

Exhibit 1: Estimated annualized burden hours

Survey Version	Number of Respondents	Number of responses per respondent	Hours per response	Total Burden hours
Medicare Disenrollee Survey, Stand Alone Prescription Drug Plan (PDP) Version	7,446	1	0.18	1,340
Medicare Disenrollee Survey, Medicare Advantage (MA-PD and MA-Only)	36,426	1	0.22	8,014
Total	43,872	1	-	9,354

Note: the number of respondents was computed as follows using data from the 2017 sample: (122,746 MA sampled * .297 response rate)= 36,426 respondents. (23,876 PDP sampled * .312 response rate)= 7,446 respondents.

Exhibit 2 shows the survey participants’ cost burden associated with their time to complete a survey. The total cost burden is estimated to be $467,325.

Exhibit 2: Estimated annualized cost burden

Survey Version	Number of Respondents	Total Burden hours	Adjusted Hourly Wage Rate*	Total Cost Burden
Medicare Disenrollee Survey, Stand Alone Prescription Drug Plan Version	7,446	1,340	$49.96	$66,946
Medicare Disenrollee Survey, Medicare Advantage (MA-PD and MA-Only	36,426	8,014	$49.96	$400,379
Total	43,872	9,354	-	$467,325

*Based upon the mean hourly wage for “All Occupations” (Occupation Code 00-000) of $24.98 per hour, as shown on the U.S. Bureau of Labor Statistics website, plus an estimate of fringe based on 100% of the mean hourly wage https://www.bls.gov/oes/current/oes_nat.htm#00-0000 (last accessed on 6/25/2019).

B13. Capitol Cost

We have no capital costs

B14. Cost to Federal Government

The total cost for design, data collection, analysis, and contract-level report production per year is $1,717,335.

B15. Changes to Burden

This request seeks approval of an estimated 9,354 hours of respondent burden per year to assess reasons for disenrollment from MA and PDP contracts. This represents a slight decrease in total burden compared with our 2017 (12,754) burden-hour estimates. This decrease in total burden hours is because we now estimate fewer respondents (based on a slightly lower response rate) and fewer hours per respondent (due to slightly shorter survey instruments than in 2017). The actual respondent burden is subject to change between years depending on shifts in number of PDP and MA contracts and/or changes in response rates from year-to-year. It is important to maintain flexibility and consider larger sample sizes that will preserve adequate contract-level reporting reliabilities in the event of increases in the number of PDP and MA contracts and/or declines in response rates.

CMS’s contractor analyzed responses to the individual disenrollment survey items to assess the number of beneficiaries screening out of the survey and reasons for screen out, and to identify potential problems with specific survey items (i.e., inappropriate skips or large fraction of missing responses). Review of the survey responses flagged a couple of items where beneficiaries sometimes inappropriately skipped questions. CMS proposes to remove the screener item to the Star Ratings questions (“Have you ever seen the Medicare Star Rating for any health plan?”), as we noted a large number of skips in this section. CMS also proposes removing the last question in this Star Ratings question series (“Did you consider the Medicare Star Ratings when trying to choose a plan?”) because it had a significant fraction of missing responses. After these removals, the Star Ratings Items on the survey would be composed of two items (“Did you leave your former plan because it got a low Medicare Star Rating?”, and “Did you leave your former plan because you found another plan with a higher Medicare Star Rating?”).

In our review of the survey responses, we also found that the items in the Spanish language version of the survey that ask about Spanish-language materials received from their former health plan also had a large fraction of missing responses (“Did you ever need written information from the plan in Spanish?” and “How often did the plan give you written information in Spanish?”). Given that these items only are used in the Spanish version of the survey and that the Spanish version of the survey is primarily used in Puerto Rico where health plans are required to provide health plan materials in Spanish, CMS decided that these items were not necessary and also proposes dropping them from the survey.

There was interest by CMS in understanding two other possible reasons for plan switching. CMS proposes adding two new items to address reasons for disenrollment related to a change from the prior year in the financial and/or health status of the individual: 1)“Did you leave your former plan because a change in your health meant the plan no longer met your needs?” and 2) “Did you leave your former plan because it turned out to be more expensive than you expected?”

Lastly, CMS proposes adding and subtracting items in response to public comments received. These proposed changes are further explained in Section B8 and the attached responses to public comments and survey item crosswalk.

In summary, the revised MA English-language version has a net reduction of 4 items (8 dropped, 4 added) and the revised PDP English-language version has a net reduction of 6 items (8 dropped, 2 added). For the Spanish-language versions of the survey, we dropped an additional 2 items. We have revised the burden estimates to reflect these revisions as discussed in Sections B12 and B15.

The proposed revisions reduce our estimated completion time per survey from 12 to 11 minutes for PDP survey responses, and from 14 to 13 minutes for MA responses.

On an annualized basis, the estimate of respondent burden is also reduced from the estimate provided in our 2017 OMB application due to a reduction in response rates from ~40% (2017 OMB application assumption) to ~30% (based on observed response rates for the survey in years 2016 and 2017).

B16. Publication/Tabulation Dates

The general schedule for publication of results from the PDP and MA plan disenrollment reasons surveys is as follows. (1) Survey fielding for the prior year’s disenrollee surveys is completed in April. (2) Data cleaning and processing is completed in May, and (3) calculation of contract-level estimates for reasons of leaving composites and single items is conducted in June/July, including weighting and case-mix adjustment). (4) The survey contractor provides CMS with contract-level scores on reasons for leaving composites that are posted to drill-downs on Medicare Plan Finder in July (and CMS posts to the plan preview page in the summer and to Medicare Plan Finder in the Fall of each year). (5) Individual contract-level reports on results from surveys from the previous year’s disenrollees are distributed to the health and prescription drug plans by e-mail and FedEx in late July/early August.

For surveys of beneficiaries who disenrolled from their contracts in calendar-year 2019 for example, the schedule proceeded as follows:

April 2020 – finished data collection for surveys from calendar year 2019 (January 2019 through December 2019)
May-June 2020 – conducted data cleaning and processing of 2019 survey results
June/July 2020 – computed weights and calculated contract-level estimates of reasons for leaving composites and single items; applied weighting and case-mix adjustment to derive estimates
July 2020 – prepared 2019 contract-level estimates of reasons for leaving composites, which CMS posted to the plan preview page later in the summer/fall)
September 2020 – distributed contract-level reports of survey results from 2019

disenrollees to the plan Medicare Compliance Officers

We anticipate a similar schedule for 2021 for processing and publishing the results of surveys of beneficiaries who disenrolled from their contracts in calendar-year 2020. This process repeats annually.

B17. Expiration Date

CMS will display the new expiration date for OMB approval of this information collection on the survey, once OMB approval has been obtained (see attachments VI, VII, and VIII) which now include text “The valid OMB control number for this information collection is 0938-1113 (Expires: TBD).”

B18. Certification Statement

There are no exceptions to the certification statement identified in item 19 of OMB Form 83-I associated with this data collection effort.

List of Attachments

Attachment I. MA-PD Prenotification Letter

Attachment II. PDP Prenotification Letter

Attachment III. MA-Only Prenotification Letter

Attachment IV. Wave 1 Cover Letter

Attachment V. Wave 2 Cover Letter

Attachment VI. MA-PD Survey

Attachment VII. Stand Alone PDP Survey

Attachment VIII. MA-Only Survey

Attachment IX. Plan Report Sample

Attachment X. Analyses Addressing 2017 OMB Questions About Year- to-Year Variability of Disenrollment Reason Scores

1 Contract-level reliability is a zero-1 index with values of .7 or higher considered adequate and value of 1 being perfect. The value indicates the proportion of variation in the reported scores that is due to true differences between contracts rather than “noise” from limited sample sizes.

2 Elliott MN, Brown J, Lehrman WG, Beckett MK, Hambarsoomian K, Giordano L, Goldstein E. 2012. A Randomized

Experiment Investigating the Suitability of Speech-Enabled IVR and Web Modes for Publicly Reported Surveys of Patients’ Experience of Hospital Care. Medical Care Research and Review 70(2): 165-84.

3 Zickuhr, K, and Madden, M. 2012. Older adults and internet use. Pew Internet & American Life Project

File Type	application/vnd.openxmlformats-officedocument.wordprocessingml.document
Author	Mitch Bryman
File Modified	0000-00-00
File Created	2021-01-14