Attach1_Memo_0925-0538_FDA_question_justification

Date: January 23, 2013

TO: Office of Management and Budget (OMB)

Through: Seleda Perryman, Report Clearance Officer, NIH

Vivian Horovitch-Kelley, NCI OMB Project Clearance Liaison, NCI

FROM: Bradford Hesse, Project Officer,

Health Information National Trends Survey 4 (HINTS 4),

Division of Cancer Control and Population Sciences,

National Cancer Institute/NIH

SUBJECT: Health Information National Trends Survey 4 (HINTS 4),

OMB No. 0925-0538, exp. 10/31/14

Per our telephone conference call on January 9, 2013, attached please find further information about the proposed HINTS-FDA instrument. As you will recall, the January conference call was in response to a memo submitted by NCI on November 27, 2012 requesting guidance from OMB about adding the FDA cycle of data collection to the HINTS submission. The memo can be found in Attachment A. During the January 9 conference call, OMB requested further information about the HINTS-FDA survey items, and specifically to understand the distinctiveness of HINTS-FDA content when compared to other surveillance sources.

Background Information

As with all cycles of HINTS data collection, approximately 50% of the instrument will be made up of standard HINTS questions on health communications and information-seeking behaviors in addition to a standardized demographics section. The other 50% of the instrument varies each cycle. We believe that this is one of the unique advantages that data collection through HINTS offers: the ability to associate varying measures of attitudes and behaviors with health communication factors that are not collected elsewhere. The same will be true of the proposed HINTS-FDA instrument. We anticipate that approximately 50% of the HINTS-FDA instrument will consist of the standard HINTS questions. The remainder of this memo will focus on the 50% of the instrument that will be new to the HINTS-FDA instrument.

The majority of the FDA portion of the HINTS instrument will be focused on questions to inform FDA’s regulatory authority on tobacco, including tobacco prevention campaign planning. Our understanding is that OMB would like further justification for these items and is not concerned about the items supplied by FDA’s Office of the Commissioner (medical products and devices) or FDA’s Center for Food Science and Nutrition (supplement labeling). Therefore, the medical product and supplement labeling questions are not addressed in this memo or the accompanying attachments.

Notes about Justification Table (Attachment B)

Attachment B provides detailed information about the tobacco questions proposed for HINTS-FDA. The spreadsheet contains:

1. Tab A, which outlines the broad constructs being explored with the HINTS-FDA instrument, and whether those constructs are assessed on other federal surveys, and

2. Tab B, which lists every tobacco item that is planned for the HINTS-FDA cycle, with a justification for why it is being asked and how it will be used in analysis, as well as whether it is duplicative of items on other federal surveys.

Of note, since the time OMB received the original memo from NCI, the HINTS-FDA instrument has been reduced quite a bit. The following topics/constructs and items have since been excluded and therefore are not included in Attachment B:

• Graphic Warning Labels section completely eliminated;

• Addiction items in tobacco use phenotypes section reduced from 22 to 4 items;

• Attitudes toward source attribution for tobacco control messaging and attitudes and beliefs about product regulation sections merged, 2 items total; and

• A number of items that were duplicative in the instrument but located in different sections attached to the original memo have been combined/eliminated (i.e., asking how many cigarettes are smoked per day was originally assessed with two questions, one in the addiction section and another in the phenotypes section, one of those items has been eliminated).

The attached spreadsheet outlines overlap between items in HINTS-FDA and other national surveys. We have indicated where those surveys have overlapping constructs/topic areas (Tab A), and then indicated overlap and analysis plan in more detail at the item level (Tab B). A HINTS-FDA item that is “Included” in another survey indicates that that item (or the item with only very minor changes, such as changing a word or the phrasing of the item) appears in another federal survey. A HINTS-FDA item that is “Similar” to an item in another survey indicates that that item appears in a highly edited or changed form, although the items measure approximately the same thing. In this case, for example, the timeframe of the item may be different (e.g., ever used X vs. used X in the last 12 months), or may ask about only a subset of tobacco products used in the HINTS-FDA item. A HINTS-FDA item that is “Indirect” indicates that that the item in the other survey could be used to determine a response to the HINTS-FDA item (e.g., if a HINTS-FDA item asks if someone has ever heard of a certain tobacco product the other survey may ask if the respondent has ever used that product, therefore we could infer that if someone has used a product they are aware of it. These “Indirect” items are not seen as being fully comparable to the HINTS-FDA items in terms of measurement objectives, however for the sake of inclusiveness we have indicated them in the spreadsheet. Among items proposed for HINTS-FDA in column M of Tab B, we have indicated items that have been previously fielded as part of HINTS “Previously included in HINTS '03, 05, 07 or '12”, items previously fielded as part of HINTS that have been slightly modified to be relevant to tobacco “Modified version of item previously included in HINTS '03, 05, 07 or '12” and those items that are new to HINTS proposed to be fielded as part of HINTS-FDA “New item proposed for HINTS-FDA”.

Similarities and Differences Between HINTS-FDA and other Surveys

Although many other national surveys use a number of the items being proposed for HINTS-FDA, there are a number of key differences to note as they relate to respondent burden.

• HINTS-FDA is only administered to adults whereas a number of other surveys (NYTS, PATH, MTF, YRBS, and NSDUH) include youth samples or are only administered to youth samples.

• The tobacco use phenotype questions section has the largest overlap with other national surveys as would be expected. Notably, this section on HINTS-FDA does not serve the same purpose as it does on other surveys. Those surveys use the information to establish prevalence estimates, track patterns of use, and monitor the impact of policies on use. In HINTS-FDA these items, of which we use a restricted set of commonly-accepted and used measures of tobacco use, serve to segment the respondents and look at associations between tobacco use and the other sections of HINTS-FDA that do not appear on other surveys, mainly media use and exposure and communication items. For example, an analysis on HINTS-FDA might explore the following research question: “Do current cigarette smokers differ in their trust of tobacco health information from various sources compared to non-smokers and/or current users of other tobacco products? How is that association influenced by health information seeking behavior about tobacco?” Although some of these items are duplicated on other national surveys, their inclusion with the other communication-specific HINTS-FDA constructs is necessary because those analyses must use constructs from the same dataset.

• Although other surveys do include a few communication questions, they mainly focus on the amount of use of various forms of media (e.g. hours spent watching TV), but do not focus on trust in health information from those sources, or other communication-related questions that are unique to HINTS..

• Similarly, for exposure questions (Tobacco Message Exposure Source/Frequency/Reaction), other surveys that include similar items tend to focus on direct promotion by tobacco companies (e.g., getting coupons in the mail), but do not go into the same level of detail that HINTS-FDA will on the type of message, the channel through which it was observed and the frequency of noticing the message. This level of detail is critical for HINTS-FDA to inform CTP’s communications efforts as CTP is actively planning a number of media campaigns and other communication efforts.

Request to OMB

We hope this response clarifies both the unique nature of the proposed HINTS-FDA instrument and the additional HINTS-FDA data collection cycle’s place within the overall HINTS program. NCI requests guidance from OMB on how to proceed in adding the HINTS-FDA cycle to the current HINTS submission. As noted in the previous memo (Attachment A), we would need to change the overall burden estimate and are unsure whether this is considered a substantive or non-substantive change request. We welcome further discussions with OMB on this matter.

Attachments:

Attachment A: Memo from NCI to OMB dated November 27, 2012

Attachment B: Justification tables for HINTS-FDA survey items

3