Supporting Statement Attachment B

B. COLLECTION OF INFORMATION EMPLOYING STATISTICAL METHODS

B.1 Respondent universe and sample

There are approximately 42 million Medicare. From this universe, we will select sample large enough to yield 3,500 completed survey from an estimated sample of 4,000. The 3,500 completed surveys in calendar 2007 will allow estimates of proportions with a 95 percent confidence interval half-width less than 2 percentage points.

The response rate for similar surveys of the Medicare population,. Such as the Medicare CAHPS survey, conducted by contractors has ranged from 75 to 80 percent. We anticipate a comparable response rate for the new survey.

B.2 Information collection procedures

The sample for the Patient Activation Survey will be drawn from MBD files. The list of beneficiaries will be sorted by Zip code and by gender within Zip code. A systematic sample (random start and a sampling interval equal to the universe size divided by the desired sample size) will then be drawn. The sample draw will be sufficient to yield the required number of interviews for the month assuming a 75 percent response rate.

The data collection contractor will mail a pre-notification letter (Attachment 1) informing sampled individuals of the survey, its voluntary nature, the legislative authority under which the survey is conducted, and Privacy Act provisions. A second mailing will include the survey questionnaire (Attachment 2). Two weeks after the initial questionnaire mailing all sampled persons will be sent a thank you/reminder post card (Attachment 1). Two weeks later, a second questionnaire will be mailed to all sampled individuals who have not yet returned a completed form. After another three weeks, interviewers will attempt to complete the survey over the telephone with continued nonrespondents. Those without telephone numbers will be sent a third questionnaire by overnight delivery.

For purposes of estimation, cross-sectional weights will be developed that reflect differential nonresponse by age, gender, and geographic region. Standard errors can be produced using software packages such as WesVarPC or SUDAAN.

B.3 Methods to maximize response rates

The methods described above have been shown to yield response rates of 75 to 80 percent with the Medicare population when the survey is of reasonable length and on a salient, non-threatening topic. The following procedures will also encourage response:

• Interviewer training will emphasize the difficulties in communicating with the older population and ways to overcome these difficulties.

• A toll-free number will be available at Abt, the contractor who will be responsible for the fielding of this survey, to answer respondents’ questions.

• Interviewers and operators answering the toll-free number will be given answers to frequently asked questions and instructions on how to address respondents’ concerns.

B.4 Tests of procedures or methods

Some of the questions are taken from previous surveys of the Medicare population, including the Medicare Current Beneficiary Survey, CAHPS and ADHCI. 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.

B.5 Statistical and questionnaire design consultants:

The contractor responsible for this survey administration is Abt Associates, a national research firm, and the responsible individuals for this survey will be:

David Newman, Ph.D, JD

Senior Associate

Abt Associates Inc.

Suite 800 North

4550 Montgomery Avenue

Bethesda, MD 20814

301-634-1749

Andrea Hassol

Survey Director

Abt Associates Inc.

55 Wheeler Street

Cambridge, MA 02138-1168

The project officer who will receive the deliverables from the supplement is:

Amy Heller

Center for Beneficiary Choice

CMS

(410) 786-9234