1-800 Medicare Survey Responses

OMB Comments On

1-800 Medicare Beneficiary Satisfaction Survey

1. Please provide some recent examples of how CMS has used and is using information from this survey.

The Vangent IVR team uses data gathered by the survey to help guide them in their recommendations/direction of their work effort.

Data gathered by the survey is, in part, used by CMS in rating Vangent during each evaluation period.

2. Were any comments received on the FR notice?

No.

3. The phone interview script promises confidentiality of responses, but the advance letter states only that no personal information will be collected, which is what is stated in A.9 of the supporting statement. SSA needs to provide the statutory authority for the confidentiality of the information in A.9 or remove this statement from the phone script.

The phone script does indeed state that the information collected will be treated as confidential. Upon direction from CMS, we can remove this from the script, or alternately, change the text to read as the advance letter does (“No personal information will be collected”)

4. Please provide more information on the sampling frame, universe size, and sample design, including what information is obtained from callers, description of any strata, as well as a statistical justification for the sample size.

Below is the data and information we have on the sampling that is done through the survey application. I’ve provided a separate attachment that speaks to the justification for the sample size.

For the Bene Sat survey,

we receive daily from the NDW a total list of callers who contacted the 1-800-MEDICARE helpline, flagged by contractor who handled the original call.  As a sample:

   The total contacts received for 6/1 - 6/7 is 30,805.

   The lowest # of contacts received on any day in that time frame:  102 on 6/5

   The highest # of contacts received on any day in that time frame: 7064 on 6/6

   The average # of contacts received: 4401

   The median # of contacts received: 5797

 

From this total list of contacts for each contractor, we select a percentage by contractor to be eligible for surveying (with a minimum selection number to be met if possible).  We currently have:

   Palmetto: 20%, minimum 100 contacts

   Vangent: 25%, minimum 100 contacts

 

We filter out or reject from the eligible survey pool those callers who don't have a phone number for callback or who have already been surveyed recently (I believe within the last month).

The information received from the NDW is the original call ID, the caller's name and address (the zip code is the only piece we particularly care about), date of the original call, caller's preferred language, the user ID of the CSR who handled the original call, the contractor who handled the original call, and the Desktop source of the contact (i.e., NGD).

 

For the Print Fulfillment Sat survey,

we receive daily from the NDW a total list of callers who contacted the 1-800-MEDICARE helpline to order a publication, flagged by contractor who handled the original call.  We also receive contacts who ordered publications online. As a sample:

   The total contacts received for 6/1 - 6/7 is 3,174.

   The lowest # of contacts received on any day in that time frame:  26 on 6/5

   The highest # of contacts received on any day in that time frame: 686 on 6/7

   The average # of contacts received: 453

   The median # of contacts received: 576

 

From the total list of contacts for each contractor, we select a percentage by contractor to be eligible for surveying (with a minimum selection number to be met if possible).  We currently have:

   Palmetto: 20%, minimum 100 contacts

   Vangent: 40%, minimum 100 contacts

   Web: 100%, minimum 100 contacts

 

We filter out or reject from the eligible survey pool those callers who don't have a phone number for callback or who have already been surveyed recently (I believe within the last month).

The information received from the NDW is the same as for the Bene Sat survey, plus the list of publications ordered by the bene.

5. (p. 3, question 15) Why is CMS collecting an additional 3,600 responses annually above the survey totals originally requested?

Other than the month of April, when we were late to change the survey application software to collect the number of surveys identified/authorized by the BCC, I wasn’t aware of our collecting more/a greater amount of surveys than had been proposed/approved.

6. How many call attempts are made to each number? What response rates were achieved the past three years?

Calls stay in the software for 3 business days , during which time we will attempt to make phone contact three times. The first attempt is made on the on day one of entry into the calling pool. If no contact is made, two additional attempts (if necessary) are then made over the next two days, after which the call is removed from the survey pool if no contact has been made.

Month-to-month response rates over the past year have been between 30% and 34%, with rare variance from this range.