Response to OMB comments

Response to OMB Questions and Action Items (CMS-10389, 0938-NEW)

Instruments

Please note: The two English versions of the survey have been integrated into a single document in revised Attachment B. Because many items are the same across the standard and alternative versions, we have instead noted only those items where alternative language and response patterns have been developed. Similar integration was also done for the Spanish versions, which have also been included in revised Attachment B, per OMB’s request.

• The respondent must be informed that the survey is being sponsored by CMS at the beginning of the survey (not just during initial contact).

This change has been made; please see revised Attachment B: Data Collection Instruments, attached.

• Unless there is specific statutory authority granted to this collection for keeping respondent information “confidential,” such assurances cannot be made.  If the Privacy Act applies, respondents should be informed that their information will be “kept Private under the Privacy Act.”  If the Privacy Act does not apply, respondents may be informed that their information will be “kept secure to the extent permitted by law.”

The language has been revised to remove the pledge of confidentiality and reference the Privacy Act instead; see revised Attachment B.

• The instruments, once approved by OMB, cannot be altered by individual vendors.  Instructions indicating that alterations are permissible must be deleted.

This change has been made; see revised Attachment B.

• Has the interview been through cognitive testing to ensure that items and response categories are optimized for the respondent population(s)?  Has the instrument been tested to ensure that the average response time is actually 20 minutes (this estimate seems to be very low given that there are close to 100 survey items)?

Yes, the instrument underwent three rounds of cognitive testing in English and one round in Spanish with the full range of disability groups included in the target respondent population. The results of this testing were used to modify items and response categories in order to facilitate comprehension and response. An expanded discussion of the cognitive testing can be found in Supporting Statement B, 4 – Testing of Procedures.

The survey introduction misstated the response time; it has been corrected to 30 minutes in revised Attachment B. In addition, Section B.12 in Supporting Statement Part A provides information as to how this estimate was derived.

• “Other” is not an approved response category when asking about race (Q #80).  This can be reserved as a response that can be recorded (if the respondent says something other than any of the 5 standard race categories) but “other” cannot be read as an option.

This change has been made; see revised Attachment B.

• Please provide the Spanish version of the survey instrument.

The Spanish version of the survey instrument has been incorporated into Attachment B; see attached.

• Will informed consent forms be used or just verbal consent?

Written consent information will be sent to all sampled participants, as described in Section 2 – Information Collection procedure of Supporting Statement Part B, along with the cover letter introducing the research project. A template for informed consent is attached (in English and Spanish). This form will be for respondents’ reference, and serve to supplement the information in the cover letter.

For phone (CATI) interviews, verbal consent will be obtained and documented. Written consent will be collected for those individuals interviewed in person (CAPI). For any sampled individual with a legal guardian, we will obtain written consent prior to contacting the program participant. These individuals will be asked to provide either verbal or written assent, depending on the data collection mode.

Please see new language in revised Attachment B.

Part A

• What is the status of IRB review at individual sites?

As of this point, CMS has not yet identified the individual testing sites. When the sites are identified, there will be coordination with the relevant IRB in each site, as per Section B.11 in Supporting Statement Part A. Should any testing site not have an IRB, the IRB at the American Institutes of Research will perform any necessary reviews. In addition, AIR’s IRB will conduct a review of all materials for each site, to assure consistency and protection for all research participants.

• A.7 indicates that respondents will be “given a pledge of confidentiality.”  Please provide that pledge as a supplemental document.

Per OMB’s directive above, we have removed reference to confidentiality in the consent materials and survey introductory script. Section B.7 in Supporting Statement Part A has been modified accordingly; see attached.

Part B

• Please confirm that the collection is not intended to be representative of HCBS in general.  Rather, with the design described, CMS intends the results to be representative of the types of HCBS represented.

Confirmed; the data collection will only be representative of the sampling frame, and the HCBS populations and services it comprises.

• Please provide a copy of the letters to potential respondents mentioned in B.2. (for both telephone and in-person respondents).

There will be a single letter to participants regardless of mode; a template for this letter is attached (in English and Spanish)

• Are there specific non-response bias analyses that are planned?

Yes; these are described in Supporting Statement Part A (Section B.2 – Information Users) and in Supporting Statement Part B (Section 4 – Non-response Analysis).

• If a CATI respondent refuses, or indicates that a face-to-face interview would be preferable would they be given the option to participate in the face-to-face interview?

Under the mode study described in Section 1 (Sampling Methods) of Supporting Statement Part B, sampled individuals will be randomly assigned to either the CATI or CAPI data collection effort. In order to maintain fidelity to the mode study methodology, individuals will not be reassigned to another mode based on preference, request or other factors (such as geographic location).