Lessons learned from the National Medicare & You Education Program.

In fall 1998 CMS implemented the National Medicare Education Program (NMEP) to educate beneficiaries about their Medicare program benefits; health plan choices; supplemental health insurance; beneficiary rights, responsibilities, and protections; and health behaviors. CMS has been monitoring the implementation of the NMEP in six case study sites as well as monitoring each of the information channels for communicating with beneficiaries. This article describes select findings from the case studies, and highlights from assessment activities related to the Medicare & You handbook, the toll-free 1-800-MEDICARE Helpline, Internet, and Regional Education About Choices in Health (REACH).

Introduction

The Balanced Budget Act (BBA) of 1997 mandated the most significant changes to Medicare since its inception. One of these changes was the expansion of health insurance options by the creation of Medicare+Choice (M+C). To support awareness of the new program changes and to help Medicare beneficiaries make more informed health care decisions, CMS initiated the NMEP called Medicare & You. The NMEP employs numerous information channels to help educate beneficiaries about making more informed decisions concerning: program benefits; health plan choices; supplemental health insurance; beneficiary rights, responsibilities, and protections; and health behaviors. The primary objectives of the education efforts are to ensure that beneficiaries receive accurate and reliable information; have the ability to access information when they need it; understand the information needed to make informed choices; and perceive the NMEP (the Federal Government and its private sector partners) as trusted and credible sources of information. There are many ways to measure success of the education program. For the program to be successful, Medicare beneficiaries must know where to go to get information when they need it and they must be aware of the basic features of the program, e.g., the existence of health plan options.

A variety of communication and information sharing vehicles are being used through the NMEP print materials; toll-free telephone services; Internet sites; a broad regional office education initiative called REACH; the national alliance network; national training and support for information givers; and enhanced beneficiary counseling and other services from State Health Insurance Assistance Programs (SHIPs). To increase the awareness of the Medicare & You information channels, an integrated campaign by the recently created Promotion and Publicity Office has begun. Other components of the NMEP are crosscutting and comprehensive assessments of the education efforts, as well as consumer testing of all publications and materials.

The NMEP is a dynamic program, and CMS is systematically studying the information needs of beneficiaries, their advocates, and its own partners to continually make enhancements and improve program efficiency. A pilot program of specific NMEP activities afforded an opportunity to study and monitor the way these specific information channels function. In 1998, two key NMEP components were implemented and tested in five pilot States (Oregon, Washington State, Arizona, Florida, and Ohio) prior to the planned fall 1999 nationwide implementation: the new Medicare & You handbook, and 1-800-MEDICARE. CMS decided to phase-in these activities in order to improve new NMEP activities through performance monitoring and assessment prior to nationwide implementation. Other monitoring mechanisms are underway to assess each of the vehicles used to communicate with Medicare beneficiaries. Since the implementation of the five-State pilot in November 1998, all NMEP activities have been aggressively monitored through mystery shopping, surveys, special research projects, focus groups, and interviews with more than 250 local officials and experts in pilot States and other communities. One key monitoring activity examines six communities to see how various NMEP components work together at the local level. As part of these case studies, a NMEP Community Monitoring Survey is administered to approximately 2,400 beneficiaries each year. The results of the monitoring activities are routinely reported to program managers to improve each communication activity.

The strategy of closely monitoring the effectiveness of different NMEP approaches to providing beneficiary assistance and incorporating lessons learned into future CMS initiatives is the key for continuing to improve efforts in helping beneficiaries to make more informed health care decisions. This article describes the primary methods of communicating with Medicare beneficiaries during the pilot and the first 2 years of national implementation, accompanied by assessment strategies, findings about the effectiveness of the information channels, improvements made to date, and the implications of the findings for the education campaign.

Medicare & You Handbook

Background

BBA 1997 mandated that general and managed care plan comparison information be mailed to all current beneficiaries by October 15 of each year, beginning in 1999. In early November 1998, CMS mailed the first Medicare & You handbook that contained local comparative information to 5.1 million Medicare beneficiaries in the five pilot States. A condensed Medicare & You bulletin that included information about the new health plan options and coverage for preventative services was mailed to beneficiaries in the remaining 45 States and territories. In 1999, CMS rolled out its new Medicare & You handbook nationally.

Copies of the 2001 handbook were mailed to approximately 34 million beneficiary households during fall 2000, and approximately 200,000 copies are mailed to new Medicare enrollees each month as they become eligible for Medicare. Through 1-800-MEDICARE, the handbook is available in English, Spanish, Braille (English only), audio, and large print, as well as a library edition. There are 26 region and State specific versions of the handbook with five States requiring more than one version. Newly included in the 2001 handbook were descriptions of patient rights and protections, preventive benefits, and availability of health care plan options including the Original fee-for-service Medicare, and managed care organizations.

Assessment Strategy

The evaluation of the handbook has consisted of (1) four different types of surveys, (2) numerous focus groups, (3) cognitive interviews, (4) expert reviews (including those by low-literacy experts), and (5) feedback postcards inserted into a sample of handbooks. As part of the consumer testing of the 2000 handbook, for example, approximately 200 cognitive interviews were conducted with beneficiaries in California, Florida, Illinois, and New Jersey during May 1999.

Findings

The percentage of beneficiaries who remember receiving the handbook has increased from 70 percent (1999 handbook) to 77 percent (2001 handbook) according to the NMEP Community Monitoring Survey (in the five original pilot States)—a statistically significant change (t=5.11, p<0.001). Survey results from February 2001 including all six sites suggest that approximately 70 percent of beneficiaries realize that the handbook is a government publication. However, 48 percent think that the document is sent by the Medicare program, 6 percent from the Social Security Administration, 17 percent from another government agency, 4 percent from an insurance company or a managed care plan and one-third do not know who sent them the handbook. Of those who do remember receiving it, approximately 76 percent had glanced through it, read parts of it, or read it thoroughly. Approximately 40 percent of managed care disenrollees, who were surveyed through the NMEP case studies and through surveys conducted in four additional sites during January and February 2001, reported that they had used the handbook to find out about their insurance options. About 13 percent of the disenrollees reported using the quality and cost comparison information provided in the handbook. A total of 5,706 beneficiaries were interviewed across the case study sites and the four additional sites impacted by managed care terminations. Out of the total sample, approximately 40 percent had been disenrolled from managed care plans as a result of Medicare plan terminations as of January 1, 2001.

Most beneficiaries both in 1999 and 2001 found the handbook to be very or fairly easy to understand (Table 1). We found that the 2001 handbook is easier for less-educated beneficiaries to read compared with the 1999 and earlier editions. Of those who did not graduate from high school, 27 percent found the 1999 edition fairly or very difficult to read, compared with only 22 percent who found the 2001 edition fairly or very difficult to read—this change is not statistically significant, however.

Table 1

Percent of Beneficiary Survey Response to the Medicare & You Handbook: 1999 and 2001

Beneficiary Response	1999	2001
	Percent
Very Easy	19	21
Fairly Easy	59	60
Fairly Difficult	13	11
Very Difficult	2	3
Refused/Don't Know	7	6

The question was asked, “How easy was the handbook to understand?”

SOURCE: Centers for Medicare & Medicaid Services, National Medicare Education Program. Community Monitoring Survey, 1999 and 2001.

The NMEP Community Monitoring Survey indicates that two-thirds of beneficiaries are satisfied with the handbook. Survey data as of February 2001, indicate that 66 percent of beneficiaries are satisfied or very satisfied, 17 percent are neither satisfied/nor dissatisfied; 4 percent are dissatisfied or very dissatisfied; and the remaining 13 percent don't know or refused to answer this question.

Improvements

All of the testing and assessment activities have consistently shown that most beneficiaries use the handbook as a reference document; and as such the Medicare & You has been revised to reflect this. The 2001 copy includes three pages summarizing key information for easy reading, highlights of important changes to Medicare, and an enhanced table of contents to make it easier for beneficiaries to find the specific information they need. Other changes have been made as a result of the assessment activities.

Simplification of Language. With feedback from consumer research, including cognitive interviews and focus groups, and through plain language reviews, CMS has continued to try to simplify the language and make it more understandable for less-educated beneficiaries.

Information on Cover. The monitoring work indicated that many beneficiaries were not aware of what the handbook was when it arrived in the mail. The 2000 and 2001 editions include information on the front cover describing what is contained in the handbook and explains that the document is published by the Federal Medicare agency.

Toll-Free Telephone Services

BBA 1997 directed that a toll-free helpline be maintained to handle inquiries about benefits and beneficiaries' available options under the M+C program. CMS phased-in the 1-800-MEDICARE helpline between November 1998 and March 1999. (The five pilot States received early access.) By March 1999, the helpline was accessible nationwide. The helpline operates on a 24-hour/7-day a week basis, and is staffed by customer service representatives (CSRs) from 8 a.m. to 4:30 p.m. local time, Monday-Friday. At other times, the helpline has a prerecorded message that assists callers requesting local Medicare health plan compare information, copies of the handbook (in English or Spanish), and answers to frequently asked questions. The contractor for the helpline is required to answer 80 percent of all calls within 30 seconds.

The CSRs help answer general Medicare questions, questions related to health plan choices, process requests for plan comparison information and plan disenrollment forms, and make referrals to other informational sources when appropriate. This service is provided to callers in either English or Spanish. Also, a TTY line (telecommunications device for the deaf and hearing impaired) is available.

More than 4.5 million telephone calls have been handled since November 2, 1998—the startup date for the five pilot States and the March 1999 nationwide implementation of the helpline. Currently, approximately 300,000 telephone calls are handled per month with annual projections for fiscal year 2001 to be 4.8 million—averaging about 5 minutes per call. The focus is on M+C options, benefits, enrollment, replacement cards, and coverage issues. When sampled beneficiaries have been asked by CSRs how they have heard of 1-800-MEDICARE, responses include: the handbook or other Medicare publications (37 percent); Social Security Administration referrals (18 percent); newspaper, radio, and other media (12 percent); health plans (6 percent); friends and family (5 percent); and other (22 percent).

In addition to examining data collected from telephone calls (including performance on 25 different indicators such as length of call and service level), CMS has contracted with Abt Associates, to conduct a customer satisfaction survey and mystery shopping. In the customer satisfaction survey, a sample of the helpline callers are recontacted and asked about the topic of the telephone call, satisfaction with the information and service provided, and demographics. A total of 3,719 telephone satisfaction surveys have been conducted from November 1998-January 2001. Mystery shoppers ask a series of questions for which specific answer comments are pre-identified and record information about the accuracy and consistency of answers, the appropriateness of referrals, and the courteousness of the CSRs.

Mystery shopping helps determine whether the CSRs can use their desktop scripts to answer questions or assist in interpreting a script to fit the question being posed. The assessment is not a test of the helpline desktop script; rather, it is a test of the CSR's ability to provide complete and accurate answers to the caller's questions. By virtue of asking the same type of question hundreds of times, the accuracy and the consistency of answers can be assessed. CMS and Abt Associates have jointly selected topics and answer components to all questions posed by the mystery shoppers. Over time, the emphasis has switched from the CSR's qualities (e.g., courtesy, opening and closing, etc.) to focusing more on the content of the answer. This is because the CSR courtesy evaluations have shown very good results (particularly during the past year) and this has improved consistently over time. Also, there are other assessment tools that assess CSR courtesy and interaction skills, i.e., CMS's quality call monitoring form and the contractors' internal quality control monitoring systems in the call centers.

Findings from the 2001 NMEP Community Monitoring Survey in the case studies suggest that 14 percent of beneficiaries call a 1-800 number (including toll-free numbers for carriers and fiscal intermediaries) to obtain information about Medicare, while 6 percent call the helpline. In general, callers were satisfied with the service provided by the CSR. Of all callers surveyed since 1998, 61 percent were very satisfied, 24 percent were satisfied, 3 percent were neither satisfied nor dissatisfied, 5 percent were dissatisfied, 5 percent were very dissatisfied, and the remaining 2 percent refused to answer this question. Overall, the majority of callers who spoke to a Medicare CSR reported that their telephone calls were handled well. About 47 percent rated the overall performance as excellent, 34 percent as very good, 13 percent as good, 4 percent as fair, and 1 percent as poor. Regarding specific characteristics, the majority rated all characteristics as either excellent or very good. Callers were the least satisfied with the thoroughness of knowledge of Medicare program information (27 percent rated this as fair or poor) and with the time it took to reach a CSR (10 percent rated this as fair or poor).

Overall, 57 percent of the callers found the automated helpline system easy to use, however, 23 percent found it confusing, 4 percent found it neither easy or not easy, and 10 percent had no comment. The most common reasons callers gave for confusion included: no option matched their concern, they wanted to speak to a real person, there were too many options, the recording went too fast, they did not understand how to use the system or what the CSRs were saying, and they did not have touch-tone service.

The callers' assessments of how well their questions were answered were positive, overall: 73 percent felt that all their questions had been answered, 19 percent felt they had some of their questions answered, and 6 percent felt that none of their questions had been answered. When asked to indicate how much trouble they had to go through to get all the information and answers they wanted, including if they were referred to another number, 41 percent said they went through no trouble at all, 24 percent went through a little trouble, 21 percent went through some trouble, and 13 percent went through a great deal of trouble.

Approximately 2,500 mystery shopping calls were conducted from December 1999 through mid-March 2001 with results indicating that customer service is consistently high, and that inaccuracies in the answers are rare (only when the CSR did not read from a script). During that time period, 91 percent of the calls were rated as understandable by the mystery shoppers. Eighty-five percent of all callers receiving a referral also received some information from the helpline. The overall courteousness of the CSRs was rated very high with 94 percent of telephone calls opening and closing courteously, and almost all of the answers to calls included correct information. Although over time the completeness of answers to questions has improved, there still remains some variation among CSRs in terms of how they have answered specific questions.

Scenario for a Mystery Shopping Question

The following is an example of one question that has been asked more than 500 times through mystery shopping : “My parents are considering joining an HMO, but lately I've been reading about HMOs changing their benefits or dropping out of the Medicare program altogether. If my parents join a plan and later it reduces benefits or leaves the Medicare program, what will happen to them?” Each time the question is asked it is phrased slightly differently. The following are some of the answer components provided, with the first three being key.

If a health maintenance organization (HMO) discontinues Medicare, you can return to Original Medicare or enroll in another Medicare HMO. (86 percent)

If an HMO discontinues Medicare, you are still covered by Medicare; there are options for the types of Medicare coverage available to you. (67 percent)

If an HMO discontinues Medicare and you return to Original Medicare you can purchase a medigap policy A, B, C, or F within 63 days after your HMO coverage ends. (20 percent)

If an HMO discontinues Medicare, you can switch to another Medicare HMO and you must be accepted (plans cannot discriminate based on prior health or service use).

During your first 12 months in a Medicare HMO, you can disenroll and repurchase your old medigap policy at the same price (if it is still being offered in your State).

You can disenroll from a Medicare HMO at any time, for any reason.

The appropriate referrals for the question are HMO, SHIP, Medicare handbook, and CMS regional office. Based on the 512 calls placed concerning this question since February 2000, there were 44 answer patterns: 1 given for 81 calls, another for 66 calls, 2 for 147 calls, and 14 calls each with a unique pattern of answers. The majority of the calls (397 or 77 percent) did not result in a referral—however, those that did receive a referral, 45 went to SHIPs, 16 went to HMOs, and 19 went to other. Abt Associates (evaluation contractor), summarized that 17 percent of the telephone calls received appropriate referrals, 79 percent were appropriately not referred, and 4 percent were not appropriately referred.

The assessment of the recent data from the helpline mystery shopping indicated that there is room for improvement in providing more complete (few telephone calls include all of the information that could be conveyed) and consistent answers. For some questions this is truer than others. Also, the assessment of helpline mystery shopping data indicated that referrals given for the same question were inconsistent, although most were appropriate. Thus, the challenge in the future is to provide more complete and consistent answers.

Mystery shopping monitoring for the past year indicates that customer service (courtesy, use of hold, waiting time, etc.) has been consistently very high and that inaccuracies are rare; most information conveyed is accurate and in most cases it is clear that the CSR is reading from the desktop script. For the rare inaccuracies in the responses, Abt Associates determined that the CSRs were not reading from the desktop script, but found that they often, and appropriately, offered to mail printed materials.

Information obtained from these assessment activities is used to continuously improve the service provided by the helpline including: streamlining the knowledge base that CSRs access to answer questions; enhancing training for the CSRs; simplifying the automated response unit and including instructions about what buttons need to be pushed in obtaining Medicare & You 2000 and future handbooks; and increasing publicity about the helpline. CMS continues to assess the helpline to make future improvements in its operations.

Internet

BBA 1997 mandated that an Internet site be developed to provide accurate and reliable information to beneficiaries on Medicare in general and plan comparison and quality information to promote informed choices. CMS launched its beneficiary user-friendly Web site (www.medicare.gov) in June 1998. Information on the site currently includes the Medicare and You 2001 handbook, a calendar of informational events and activities, and lists of resources and telephone numbers for obtaining additional information about Medicare. The Web site also hosts several comparative databases: Medicare Health Plan Compare; Medigap Compare; Nursing Home Compare; and Dialysis Facility Compare.

Medicare Health Plan Compare—contains detailed comparisons of benefits, costs, and selected results from consumer satisfaction surveys and standardized performance measurement systems of available managed care plans across the country.

Medigap Compare—provides general information about medigap insurance and aids in finding insurance companies that sell Medicare supplemental insurance plans.

Nursing Home Compare—includes detailed information about individual nursing homes around the country.

Dialysis Facility Compare—aids in locating and comparing dialysis services with respect to basic hours of operation, and clinical outcome measures such as adequacy of hemodialysis, anemia management, and patient survival.

The new prescription drug module, Prescription Drug Assistance Program, offers information on programs that offer discounts or free medications to individuals in need.

Beneficiaries and those acting on their behalf are increasing their use of www.medicare.gov. As of January 2001, site statistics logged 2.8 million page views compared with January 2000 when there were 1.5 million page views. The www.medicare.gov site received 7,726,364 page views (20,269,022 hits) for the first quarter of calendar year 2001. In comparison, for the same time period, the Web sites for the National Women's Health Center (www.4woman.gov), Office of Women's Health, logged 3,123,003 page views (13,865,618 hits), and the Office of Public Health Service (www.healthfinder.gov) Office of Disease Prevention and Health Promotion logged 4,390,537 page views (18,225,233 hits).

Continuous assessment activities include automated tracking of utilization, online users bounceback surveys, computer lab sessions followed by focus groups, and an expert review carried out by Web site designers and two consultants familiar with subject areas on the Web site. The methodology behind the bounceback form is to allow Web site visitors the option to complete an online survey. There were no restrictions preventing visitors to the Web site from completing the survey multiple times because users visit sites numerous times for various reasons. Additionally, the bounceback survey is optional so it does not represent all visitors to the site.

Since its inception in April 1999-mid-March 2001, there have been 16,693 bounceback forms filled out. To date, the largest group of respondents to the bounceback survey are current or soon-to-be Medicare beneficiaries (38 percent), followed by beneficiaries' relatives or friends (27 percent), and health professionals such as social workers or nurses (25 percent). Beneficiaries learned about www.medicare.gov from the print materials they received (especially the Medicare & You handbook) and most discovered the site to be user friendly. Approximately 85 percent of users filling out the bounceback form indicate that the site contains useful information, and approximately 88 percent indicate that the site is easy to use. Visually impaired focus group participants found the site is well designed for use with assistive technology. Overall, the expert reviewers judged the content favorably; still, recommendations to modify the presentation of the information in a more consumer-oriented manner were offered. In their assessment of the site in 1999, the Web site design experts described the navigational features of the site to be slow and unclear.

Beneficiary use of the Internet continues to grow—according to data from rounds 18, 24, and 27 of the Medicare Current Beneficiary Survey. The percentage of beneficiaries with access increased from 6.8 percent in 1997 to 21.3 percent in 1999 to 28.5 percent in 2000 (Centers for Medicare & Medicaid Services, 1997, 1999, 2000). Data from the NMEP Community Monitoring Survey indicate that at the study sites, the percentage of aged beneficiaries using the Internet to seek Medicare information increased from 1.4 percent in October 1998 to 3.0 percent in February 2000 and remains essentially the same in February 2001. Based on a multivariate analysis, results show that among aged beneficiaries, Internet usage to seek Medicare information is negatively associated with being older, female, not as educated, and having less knowledge of the program. These statistically significant results indicate the need to have in place other communication vehicles to provide for those audiences not using the Internet when seeking information about Medicare.

A number of improvements have been made based on feedback received from users of www.medicare.gov and the expert reviewers. Participants in the computer lab sessions and focus groups did not notice the search engine and had problems finding some of the information available, for example, telephone numbers. Recommendations were made by the participants about keeping the use of graphics to a minimum for faster loading whereas, the expert reviewers suggested less use of jargon, bureaucratic language and acronyms, and more description about specific links such as “Wellness” and “Medicare Compare.” Some improvements that have been made based on their comments include: simplifying language; enhancing searching capacity; and making the publication section more user-friendly for novice users or those using assistive technology. Additionally, recent changes have been made including the appearance, navigation, and feel of the site to make it more user-friendly in general and especially for those visually impaired; and making printing from the site easier. The feedback from the focus group with Medicare beneficiaries who were under age 65 cited the lack of information about kidney disease, the Web site now has a section called Dialysis Facility Compare.

Reach

BBA 1997 required that annually during the month of November, in conjunction with the annual coordinated election period, that a nationally coordinated educational and publicity campaign provide M+C eligible individuals with information about health care plans and the election process. In compliance with BBA 1997, each of CMS's10 regional offices (ROs) since 1998 has conducted educational and outreach efforts at the regional, State, and local levels. Many of these activities are targeted to meet the specific needs of underserved populations: people with Medicare who encounter particular barriers to accessing information, due to language, location, culture, literacy, disability, etc.

The 1998 campaign activities were quite diverse across the regions and the States. As a result, in 1999 and 2000 all regions implemented core activities through the REACH campaign including: the media, health fairs, educational seminars and public presentations, local district congressional and legislative updates, expanding and enhancing partnerships, and distribution of materials. CMS central and ROs developed the REACH National and Regional Office Business Plans in 1999 to foster common goals and strategies for the outreach campaign.

In 2000, the REACH campaign that ran from August-December involved an estimated 3,247 outreach and awareness activities across the country (65 percent of these activities took place during the months of September and October). The majority of the activities during the period from August-December 2000, were public presentations that focused on Medicare health plan non-renewals totaling roughly 1,140, or one-third of the total activities for the REACH campaign last year. Many of the REACH activities were public presentations, meetings, and media intervention strategies as shown in Table 2. There was a marked increase in the total number of non-renewal events in 2000 compared with 1999, which were 640 for the same time period.

Table 2

Number of Estimated Activities for the 2000 REACH Campaign

Activity	Number of Events
Total	3,247
Educational Seminars and Public Presentations	1,629
Health Fairs	410
Public and Partner Meetings	573
Media Intervention Strategies	206
Training	172
Other	257

Outreach and awareness activities cover the period from August-December, 2000 with most activities taking place during the months of September and October.

NOTE: REACH is Regional Education About Choices in Health.

SOURCE: REACH National and Regional Business Plans, 2001.

Beginning in 1999, all regions have used nationally developed materials and the activities have been supported and assessed through single national contractors, rather than region-specific contractors. In 1999 and 2000, a wealth of data on the REACH campaigns was collected through direct observation of activities, interviews using structured protocols, and interviews with attendees of events, in order to establish a baseline for performance measurement and to analyze activities and best practices. In REACH 2001, the campaign will emphasize the use of social marketing methods by further identifying and targeting reactive and passive audiences and underserved communities.

Abt Associates' evaluation of REACH in both 1999 and 2000, concluded that conducting live events (such as health fairs and public presentations) was not a cost-effective strategy for mass education about Medicare, but they can be valuable in addressing special needs of target populations and in special contexts (such as plan non-renewals). It was further recommended that social marketing techniques be employed to determine situations in which beneficiaries are likely to be especially receptive to information. In interviews with CMS's ROs, Abt Associates found that the ROs viewed mass media as the best way to reach the large majority of beneficiaries who could be considered passive and reactive information seekers and that the media is far more cost effective than health fairs or presentations. Passive beneficiaries may be at risk for making poor decisions because they lack comprehensive information. An effective message for passive beneficiaries is: “There are Medicare information channels that you can use.” To reach passive audiences, a recommended communication activity is a mass media (radio, print, or television) campaign that focuses on communicating about the Medicare information vehicles. Somewhat different are the reactive beneficiaries, those who only seek information when a need arises, e.g., plan non-renewals. Reactive beneficiaries prefer to get information from a single, knowledgeable source, such as the handbook, helpline, Internet, or through one-on-one assistance with written materials. A recommended message for reactive beneficiaries is: “You will find answers to your Medicare questions by using the Medicare information channels.”

There were several positive findings about REACH 2000 partnering activities: one region was cited as undertaking a best practice in their secondary partner training activity, an annual program of 15 to 18 regionwide events aimed at disseminating information to about 50 secondary partners in each location. (The same region also undertook their own assessment of the flow down effect of their secondary partner training). From its assessment activities Abt Associates found that the regions concurred that there were significant improvements in regularizing partnership communications in 2000 compared with 1999. The most notable finding on general partnering strategy was the great variation across the regions, and across States within regions. The strategy tends to be highly State-specific and variations are due to the different needs in local markets, the varying interest and abilities of the partners, and the varying interests and abilities of the REACH staff on partnering.

The overall approach in REACH 2001 is to have a “nationally coordinated campaign” aimed at a higher level by continuing an emphasis on planning and strategy. The ROs continue their accountability for implementing activities and leveraging partners, and the Central Office continues to provide resources, national materials, assessment, and overall campaign coordination. Social marketing will increase in importance with the implementation of the campaign, by further identifying and targeting reactive and passive audiences and underserved communities. Quantitative criteria and social marketing research also serve as a basis for selecting campaign activities and for expenditures. As the ROs reduce face-to-face events, live presentations, and health fairs, they will continue their role for responding to crises, non-renewals, and the needs of underserved or special populations.

Targeted and Comprehensive Assessment of Education Efforts

A key monitoring activity examines six communities (Dayton, Ohio; Eugene, Oregon; Olympia, Washington; Sarasota, Florida; Springfield, Massachusetts; and Tucson, Arizona) to see how various NMEP components work together at the local level. In selecting these sites, CMS chose communities based on variations in managed care penetration rates, large employer groups, rural and urban locations, and bilingual populations. CMS continues to study these communities, chronicling relevant changes and events within each—as well as the reactions of beneficiaries and related organizations to the information NMEP provides. The case studies have identified what seems to be working in terms of the NMEP, and provided guidance on future improvements.

The community case studies include interviews and focus groups with key information providers, as well as surveys and focus groups with beneficiaries living in these communities, and data from CMS administrative files. NMEP Community Monitoring Surveys have been conducted with beneficiaries in the case study sites in October 1998, January/February 1999, January/February 2000, and January/February 2001. For each round of the survey at least 2,400 beneficiaries were interviewed—2,986 beneficiaries were interviewed in January/February 2001.

The survey measures beneficiaries' activities in gathering information regarding the Medicare program and include questions on demographics, health status, and Medicare knowledge. The surveys also provide data about situations that may affect beneficiaries' probability of searching for information—e.g., changes in health status and retiree benefits, death of a spouse, and being involuntarily disenrolled from a health plan. These questions refer to the 12-month period prior to the survey. The data collected from the case study sites cannot be generalized to the overall Medicare population; however, results from similar questions about beneficiary knowledge and use of the handbook, for example, on this survey are similar to the national Medicare Current Beneficiary Survey.

From 1998-2001, substantial managed care market changes occurred in the case study sites, with implications for the demand and supply of Medicare information. These changes included fewer available Medicare managed care plans, reductions in benefits, increases in premiums, and disruptions in provider networks. Table 3 contains information about managed care penetration and the number of managed care plans available during this period.

Table 3

Medicare Managed Care Enrollment and Number of Plans, by Study Site: 1998 and 2001

Study Sites	Percent of Managed Care Penetration			Number of Medicare Plans
	1998	2000	2001	1998	2000	2001
Dayton, Ohio	16.8	16.5	15.2	4	3	2
Eugene, Oregon	45.9	45.0	41.7	4	3	3
Olympia, Washington	37.2	41.4	38.8	4	4	3
Sarasota, Florida	12.1	11.3	NA	4	2	0
Springfield, Massachusetts	21.3	21.6	21.3	6	3	3
Tucson, Arizona	48.8	49.1	45.2	7	4	2

NOTE: NA is not applicable.

SOURCE: Centers for Medicare & Medicaid Services, Group Plan Files, 2001.

Over the 3-year study period, the number of Medicare managed care plans offered in the study sites dropped from 29 to 13, with plan withdrawals occurring at all sites. In Sarasota all four of the plans that had been present in 1998 had withdrawn from Medicare by the end of 2000, leaving beneficiaries with no option to join a managed care plan. Additionally, access to managed care plans became more difficult in two sites: a remaining Tucson plan closed enrollment into one of its plans and an Olympia plan instituted capacity limits, thereby potentially limiting enrollment. All plans in the study sites reduced their benefits and/or increased their charges for 2001. Plans at four of the study sites experienced network disruptions as providers terminated or refused to contract with the plans.

Plan terminations and major benefit or network changes were the main factors causing beneficiaries to switch plans or return to Original Medicare in the study sites. A cohort of aged beneficiaries who have continuously lived in each site since July 1998 has been tracked. In Olympia (86 percent) and Eugene (88 percent) (the most stable sites in terms of managed care offerings), of beneficiaries who were in a managed care plan at the beginning of the study in July 1998 were still in the same plan as of February 2001. The experiences of many beneficiaries in Tucson (45 percent), Dayton (56 percent) and Springfield (68 percent) have been quite different—those who were enrolled in a plan at the beginning of the study were still in the same plan in February 2001.

As of February 2001, more beneficiaries in the cohort had left managed care for Original Medicare than had switched from Original Medicare to a managed care plan. From July 1998 to February 2000, 8.4 percent of beneficiaries with managed care experience switched to Original Medicare, while over the following year 12.2 percent switched to Original Medicare. On the other hand, from July 1998 to February 2000 4.2 percent of beneficiaries who were enrolled in Original Medicare changed to managed care, whereas only 1.0 percent changed to managed care during the following year.

The majority of beneficiaries who sought information about managed care did not use CMS information sources. About 16 percent of beneficiaries in the 2001 NMEP Community Monitoring Survey said they had sought managed care information, primarily from managed care plans. Beneficiaries who sought managed care information through direct contacts cited their physicians and office staff, and friends or family members as the next most frequent sources of managed care information after managed care plans. (Few reported speaking with other sources.) However, senior centers, SHIPs, and Medicare were cited more often as sources for printed information about managed care compared with physicians and friends and family. Use of information from CMS, (helplines, printed materials, etc.) by beneficiaries seeking information about managed care increased from 7 percent during 1998 to 12 percent during 2000.

There is evidence from the four waves of surveys that many beneficiaries lack a basic understanding of the Medicare program. In the 1999 and 2000 surveys, 18 percent and 17 percent of beneficiaries surveyed, respectively, were not familiar with the terms “managed care plan, HMO, or health maintenance organization” although managed care existed in all of these communities. The percentage of beneficiaries knowing that Medicare does not cover all health care costs has increased from 85 percent in October 1998 to 89 percent in February 2000 and was basically unchanged in February 2001. Fewer than one-half of beneficiaries correctly answered questions about whether joining an HMO meant leaving Medicare. With one significant exception, beneficiaries' knowledge was unchanged from 2000 to 2001. The only area of knowledge that changed was the response to the question, “True or False: Medicare managed care plans are allowed to raise their fees or change their benefits every year?” Forty-two percent of beneficiaries answered correctly in 2001 compared with 36 percent in 2000.

The case studies have also provided data about beneficiary information seeking behavior that CMS is taking into account as the NMEP evolves. From focus groups with beneficiaries, as well as expert interviews with counselors and organizations helping beneficiaries, it is clear there is not a significant demand for general Medicare information from beneficiaries. The case studies have also indicated that most beneficiaries seek information about Medicare on an as needed basis. In the 2001 NMEP Survey conducted in the study sites, for example, rates of seeking Medicare information were approximately 77 percent for new enrollees, 89 percent for aged beneficiaries who were involuntarily disenrolled from their health plans, 77 percent for aged beneficiaries whose employer changed retiree benefits, and 73 percent for aged beneficiaries who reported a decline in health. These rates are all higher than the 66 percent rate of Medicare information seeking for the whole beneficiary survey population.

Particular characteristics seem to relate to information seeking. Based on the 2001 survey data for aged beneficiaries, the likelihood of seeking Medicare information in the study sites is positively related to being younger, having more than a high school education, having a decline in health, and having a change in their insurance in the past year. Overall, and for almost every specific type of information mechanism, there is a strong pattern showing that persons more knowledgeable about Medicare (based on a set of knowledge questions asked on the survey) are the highest users of information. More knowledgeable persons may choose to obtain more information, or persons accessing more information may become more knowledgeable.

Use of formal and informal sources of information differ by subgroups. For example, beneficiaries living alone are, not surprisingly, more likely to use providers as a source and less likely to use family and friends or plan representatives. Non-white beneficiaries are more likely to use family and friends than white beneficiaries, and far less likely to use plan representatives.

The case studies serve as a vehicle for CMS to monitor the results of the various NMEP information mechanisms at the local level. They have identified that situations and events have an important effect on increasing demand for Medicare information, and that much of the work of providing information is a local matter, with heavy dependence on intermediaries, or partners, to service it. They also have provided a deeper understanding of the complexities of informing and educating beneficiaries. To date, it is clear from the case studies that in the future CMS needs to increase targeting of information to beneficiaries who are experiencing a particular situation as well as increase beneficiaries awareness of where they can go to get Medicare information when needed.

Conclusion

CMS's monitoring efforts undertaken over the last 3 years have been intensive, which is appropriate for the scale and uniqueness of the NMEP. The monitoring work has contributed formatively to the evolution of the NMEP, though the basic structure of the approaches (mechanisms) remains as specified by Congress.

The monitoring work to date has indicated that most beneficiaries use some Medicare information, and the use of it rose sharply with the mailing of the handbook in 1998. Some modest trends in the extent of information usage have also been detected since then, particularly in the usage of the Internet and the helpline. We know, however, that unmet needs for information still exist as evidenced by: knowledge limitations; lower levels of use for older, those less educated, and possibly other beneficiary segments; and lower satisfaction with information by these and other vulnerable groups. The monitoring work also indicated that personal situations and events such as plan terminations and failing health have an important effect on increasing demand for Medicare information, and that much of the work of providing information is a local matter, with heavy dependence on information intermediaries or partners to service these needs.

CMS's monitoring activities have produced feedback on the progress of the NMEP efforts and on potential areas for improvement. Over the 3-year period, the activities have provided a deeper understanding of the complexities of informing and educating beneficiaries. CMS has used this information to change approaches as the NMEP matures—by shifting from a broad-based educational campaign to a more focused campaign of ensuring that information is available when needed and that beneficiaries and those acting on their behalf are aware of where to access Medicare information. The role of the handbook has shifted to a reference document; the use of health fairs and other public presentations has become more targeted; and increased attention is being paid to the media. The role of information intermediaries—contractors, grantees, and other organizations that provide Medicare information—is now more fully understood: such entities have an important role in the local targeting and delivery of Medicare information. Accordingly, CMS has increased its efforts to support these entities, by developing and improving their Web site, partner toolkits, and numerous documents and slides intended specifically for partner use. External events also affected the direction of NMEP, the most noteworthy being the need to direct resources for information support about managed care plan terminations from the Medicare program.

Changes have occurred not only in the communications mechanisms but in the monitoring strategies themselves. Based on early monitoring results, questions were added to the NMEP Community Monitoring Survey to enable the study of situational segments of beneficiaries (those experiencing plan terminations, changes in retiree benefits, death of a spouse, etc.) and to measure awareness of new program offerings such as the private fee-for-service option. As the NMEP increased its efforts to reach vulnerable populations, new sampling strata were added to the survey. Over the past 3 years, the emphasis of mystery shopping the helpline has increasingly focused on the content of the answers and the appropriateness of referrals.

The NMEP will continue to evolve to address the needs of the Medicare population. In fall 2001 CMS will launch a large-scale national media campaign to advertise the Medicare information resources available to beneficiaries as well as their health plan options; 1-800-MEDICARE will expand its operations by allowing a caller to speak to a live CSR 24 hours a day, 7 days a week; and www.medicare.gov will add a tool to calculate out-of-pocket costs for different health plan options based on a beneficiary's age and health status. All of the changes are being made to better address the needs of the growing and diverse Medicare population.