1. Potential Respondent Universe
The target population for the New Enrollee survey consists of all projected new Medicare beneficiaries, almost 50 million, in the continental United States (fifty states plus the District of Columbia). We are planning on obtaining 1,200 completed interviews annually.
Records of all Medicare beneficiaries are maintained by CMS.
2. Procedures for Collecting Information
Sample: As previously described, the target population for the New Enrollee survey consists of all new Medicare beneficiaries in the 50 continental United States and the District of Columbia. In order to get a representative and reliable sample of Medicare beneficiaries a sample of approximately 30,000 beneficiaries will be drawn.
Based on experiences with previous surveys, we expect a 50 percent telephone match (that is, 15,000 cases). We plan to interview 8 percent of the eligible cases, that is, in total, approximately 1,200 beneficiary interviews. CMS will implement the following steps in selecting a sample of Medicare beneficiaries:
Remove ineligible records including: beneficiaries in Puerto Rico and U.S. territories; deceased beneficiaries; and beneficiaries no longer eligible for Medicare.
Sort the administrative file by key demographic variables (geographic region, race, Medicaid eligibility, gender, and age) to help ensure that an appropriate proportion of each demographic group will be represented.
Select systematically with equal probability a sample of approximately 30,000 cases while the records are in the above sort order.
Once the sample has been selected, addresses for the sampled beneficiaries will be sent for telephone matching. To maximize the match for telephone numbers, two telephone matching services will be used. Based on the hit rate from the previous surveys, we expect to match 50% of beneficiary telephone numbers.
Target number of completes at target response rate will yield data which will provide stable national estimates for all Medicare beneficiaries. The maximum percent error for estimates of percentages obtained from a simple random sample yielding 7,000 completed interviews will not exceed 2 percent, 95 percent of the time. For example, suppose 50 percent of Medicare beneficiaries report being enrolled in managed care plans. We can be 95 percent confident that between 48 and 52 percent of the beneficiaries are enrolled managed care plans. The percent error is the largest for the 50 percent proportion and decreases as proportion moves further away from the 50 percent/50 percent split. No unusual problems are foreseen in the sampling or data collection for the New Enrollees survey activities.
Data Collection: We will use a questionnaire which has been used in previous years and modified based on findings from behavior coding over two administration cycles. The questionnaire will be modified to measure knowledge and understanding of the Medicare program. The questionnaire will be approximately 15 minutes in length.
Data will be collected using computer assisted telephone interviewing. Interviewers will be rained specifically on the administration of this study and will be monitored throughout the data collection period to ensure data collected is of the highest quality. Respondents will be sent a pre-notification letter to inform them that they will be called by a survey interviewer and to answer some of their questions in advance. The data collection contractor will mail a pre-notification letter (Appendix A) informing sampled individuals of the survey, its voluntary nature, the legislative authority under which the survey is conducted, and Privacy Act provisions.
Over the course of the project, interviewers will call Medicare beneficiaries and complete the various surveys in either English or Spanish, at the request of the respondent. Data cleaning will take place throughout the data collection period with a final round of data cleaning to occur after the last interviews have been completed. The final clean data will be weighted to account for sample selection and non-response.
There will be one national phone survey fielded, on a rolling basis for two weeks a year for three years, with a total of 1,200 respondents. The survey will be fielded in both English and Spanish.
3. Methods Used to Maximize Response Rates
To maximize response rates CMS is using a computer assisted telephone interviewing (CATI) mode of data collection.
To achieve the highest possible response rate to the survey, CMS’s survey contractor will:
Conduct interviews in English and Spanish. Although the number of the sampled respondents who will complete the interview in Spanish is estimated to be small, it is important to ensure that the opinions and experiences of this population are represented.
Use assisted proxies to help complete interviews with respondents who have hearing or speech impairments. The number of cases to be completed with assisted proxies will be very small, but these beneficiaries are likely to be users of more Medicare services so their input is critical to this task.
Send pre-notification letters to selected respondents. This measure has been shown to increase response rates and also serve as notification to respondents of an up-coming call.
Focus much attention on training interviewers to answer respondent questions. Specific training modules will be designed to prepare interviewers to communicate effectively with seniors and alleviate any concerns respondents may have regarding participation in the study and their Medicare benefits.
Have interviewers available during a wide range of times complete interviews with respondents. Since we will be conducting interviews across the United States, we will have interviewers available from 9am to midnight Eastern time, Monday through Friday, 10am to 9pm Eastern time on Saturday, and noon to 11pm Eastern time on Sunday.
Vary calling times to respondents. In an effort to contact respondents at a time that is convenient for them, the CATI scheduler will track calls to each case and insure that calls are placed to beneficiaries at different times of the day and different days of the week. When a specific appointment time is requested, the scheduler will deliver the case to an interviewer at the appointed time.
Maintain a toll-free telephone number for respondents to call with questions about the study. The toll-free telephone will be answered by a specially trained interviewer during all hours of data collection. During the hours when data collection is not underway, callers will be prompted to leave a message to have their call returned.
After data collection is complete, the data will be weighted to adjust for sampling probability and the effects of non-response. The adjustment steps will include:
Weighting to account for selection of the sample records from the
selected administrative file;
Weighting adjustment to make sampled weights match the number of
eligible beneficiaries;
Weighting adjustment to compensate for non-response in the interview
phase.
New Enrollee data was weighted using the protocol described above.
4. Tests of Procedures or Methods
Some of the questions in the New Enrollee survey are taken from previous surveys of the Medicare population, including the MCBS and the NMEP. New questions have been tested using cognitive methods with small samples of new beneficiaries. The data collection procedures have been well-tested on several surveys of Medicare beneficiaries.
To analyze the results of the New Enrollee survey, the contractor will review responses from interview sessions to identify trends and patterns in response. In reports to CMS, the contractor will highlight themes and topics where consistent responses emerge and note areas where respondents were in disagreement. The contractor will then generate data tables using SPSS, which will display data frequencies and statistical output, such as means, standard deviations, variances, etc. The contractor will provide to CMS, breakdowns of the data by wave (year one vs. year two, etc.) and relevant respondent subgroups (e.g. geographic region, demographics, retirement plan behaviors, etc.). All tabulations will show statistically significant differences by sub-group at the 95% confidence level.
5. Contacts for the New Enrollee survey
Reneé Clarke
Centers for Medicare & Medicaid Services
CMS/OEA
Mail-Stop: S1-13-05
7500 Security Blvd.
Baltimore, MD 21244
410-486-0006
Frank Funderburk
Centers for Medicare & Medicaid Services
CMS/OEA
Mail-Stop: S1-13-05
7500 Security Blvd.
Baltimore, MD 21244
410-486-1820
| File Type | application/msword |
| File Title | Supporting Statement for Paperwork Reduction Act Submission for |
| Author | CMS/OEA/CSG/DoR |
| Last Modified By | CMS |
| File Modified | 2009-05-27 |
| File Created | 2009-05-27 |