Supporting Statement Part B

Information Collection Request

New

Reaching Minority Men Where They Are: Formative Research to Build Capacity for Enrollment in Diabetes Prevention and Management Programs

GENIC submitted under generic clearance 0920-1154

Supporting Statement: Part B

Program Official/Contact

 Andrew Lanza, MPH, MSW, NCCDPHP

Division of Diabetes Translation

National Center for Chronic Disease Prevention and Health Promotion

Centers for Disease Control and Prevention

Atlanta, Georgia 30341

404-610-1294

FAX: 770-488-5939

Email: [email protected]

TABLE OF CONTENTS

Statement B

Section 1 – Respondent Universe and Sampling Methods

Section 2 – Procedures for the collection of information

Section 3 – Methods to maximize response rates and deal with no response

Section 4 – Test of procedures or methods to be undertaken

Section 5 – Individuals consulted on statistical aspects and

individuals collecting and/or analyzing data

Tables

Table 1 - Opt-in Panel Quota Sample Design

Table 2 – Focus Group Design

Table 3 – Individuals Consulted on Statistical Aspects and Individuals Collecting and/or Analyzing Data

Supporting Statement B:

1. Respondent universe and sampling methods

The respondent universe includes all minority males who are likely at risk (identified through established risk factors) for type 2 diabetes or diagnosed with type 2 diabetes. Their responses, in both the survey and focus groups, will inform key principles and elements that will be helpful for program delivery sites (for both type 2 diabetes prevention and diabetes self-management) to engage and retain more minority men. Racial and ethnic minorities consist of American Indians and Alaska Natives, black or African Americans, Hispanics or Latinos, and Asian Americans, Native Hawaiians, and other Pacific Islanders.

Survey

The sample of respondents for the survey will be drawn from Dynata’s online panel (see https://www.surveysampling.com/audiences/consumer-online/ for more information about this panel). Although the sample will be a convenience sample, the panel, which includes more than 60 million people globally, includes people from all age groups, genders, regions, education and income levels. The findings from this proposed research project are not expected to be widely generalizable to the universe of minority men with diabetes or at risk for type 2 diabetes in the U.S. Since this project is part of formative work for a pilot program, these methods are not intended to generate nationally representative samples or precise estimates of population parameters (see graphic). However, the design allows for comparisons of responses between segments of the population that will provide valuable information on how perceptions of key elements and principles are the same or different between different groups.

CDC proposes to conduct a non-probability quota sample of 1,500 minority men from the opt-in panel. Quotas were established for race/ethnicity and disease status as shown in table 1 below.

Table 1 Opt-in Panel Quota Sample Design

	Diagnosed with type 2 diabetes	At risk for type 2 diabetes	TOTAL
Hispanic	n = 300	n = 300	n = 600
Non-Hispanic Black	n = 300	n = 300	n = 600
Asian & Other	n = 150	n = 150	n = 300
TOTAL	n = 750	n = 750	n = 1,500

This sample design will allow for detection of statistically significant differences of 10% or more between subpopulations with a statistical power of between 55% and 80% at a 90% confidence level depending on the comparison being made. Comparisons made between groups with a sample size of 300 will have a statistical power of 80% whereas comparisons made between groups with sample size of 150 will have a power of 55%. The reason for this difference is that the researchers on this project have chosen to prioritize comparisons between and among the racial/ethnic groups that represent the largest total population in the U.S. over racial and ethnic groups with smaller populations given the limited resources available.

To achieve this sample size, we anticipate screening 3,000 members of the opt-in panel based on estimates of need from the panel vendor. It is anticipated that approximately half (50%) of respondents will be deemed ineligible due to not meeting the inclusion criteria.

In addition to the opt-in panel survey, CDC proposes partnering with a community-based minority health organization to collect an additional 500 responses to the survey (using the same survey instrument) in order to contribute different points of view about health behaviors than those offered by men that have opted into a panel. These data will be held and analyzed separately from the opt-in panel survey data because, given the anonymity of respondents, we will not be able to ensure there is no duplication between respondents to the opt-in panel survey and respondents to the community partners survey.

Focus Groups

CDC is also proposing to conduct six 90-minute focus groups with minority men who have been diagnosed with diabetes or are at risk for type 2 diabetes. The focus groups will be conducted in three distinct geographic locations and will ideally include 6-8 participants each but could include as many as 10 participants each for a total of between 36 and 60 participants as displayed in Table 2. Two focus groups will be held at each location – one with participants that have been diagnosed with diabetes and the other with participants that are at risk for type 2 diabetes. Both focus groups at each location will have respondents from the same race/ethnicity group and will be moderated by a skilled individual. Efforts will be made to use a moderator who is of the same racial/ethnic group as the participants. While not all race/ethnicity groups will be represented, researchers have chosen to prioritize the race/ethnicity categories that represent the largest portion of the US population (Hispanic, Non-Hispanic Black/African American) and the group with the highest prevalence of disease (Native American/Alaska Native).

Table 2 Focus Group Design

	Population Addressed
	Native American (Phoenix, AZ)	Black/African American (Charlotte, NC)	Hispanic (Miami/Tampa, FL)	TOTAL
Diagnosed with Type 2 Diabetes	6-10 participants	6-10 participants	6-10 participants	18-30 participants
At-risk for Type 2 Diabetes	6-10 participants	6-10 participants	6-10 participants	18-30 participants
TOTAL	12-20 participants	12-20 participants	12-20 participants	36-60 participants
Note: On average we expect about 8 participants per group.

Locations for the focus groups were chosen due to their high concentration of the racial/ethnic minorities that we are targeting. The locations are: Phoenix, Arizona; Miami or Tampa, FL; and Charlotte, NC. San Antonio, TX will be used as an alternate location should we need to change locations for some reason.

Focus group participants will be recruited through an established vendor in each location based on screening criteria provided by CDC. To achieve this sample size, we anticipate screening up to 300 members of the vendor’s panel based on estimates of need from the panel vendor. It is anticipated that approximately half of respondents will be deemed ineligible due to not meeting the inclusion criteria.

Since this project is part of formative work for a pilot program, these methods are not intended to generate nationally representative samples or precise estimates of population parameters. However, the design allows for collecting more rich, qualitative data than can be collected on the national survey and is expected to provide information on the perceptions of key elements and principles identified in the landscape assessment.

2. Procedures for the collection of information

Survey

The data collection subcontractor, Leavitt Partners, will be responsible for coordinating data collection activities, collecting and summarizing information, and preparing topline reports. The survey will be administered through Qualtrics, an online survey platform, and survey invitations will be sent by Dynata, an online panel provider.

Recruitment and screening

Opt-in panel survey respondents will be recruited through the Dynata online panel. Dynata has profiled their panels on a number of different characteristics and will use that information to target respondents who are likely qualified to take the survey. These individuals will be sent an email invitation (attached) to participate in a survey. When the individuals click on the “start survey” button, they will be routed to a screening survey that screens them for participation for up to ten different surveys. Dyanata’s proprietary Dynamix system uses the responses to the screening survey to route the individuals to one of the surveys that they qualify for. Dynata has shown that this methodology removes self-selection bias and minimizes drop rates since participants don’t have the option to stop a particular survey and switch to a different one.

Community partner survey participants will be recruited by the community minority health organizations. We envision a link to the survey will be sent out through newsletter distribution and posted on social media accounts for the community partner organization.

Eligibility criteria for the survey are as follows:

• Male, and

• Minority race/ethnicity, and

• Aged 18+

• Disease status

  • Diagnosed with diabetes or

  • Diagnosed with prediabetes, or

  • Diagnosed with hypertension, or

  • Has a family history of type 2 diabetes (parent or sibling), or

  • Is physically inactive, or

  • Aged 45 or older, or

  • BMI >23 for Asian Americans, BMI > 25 for all other races

If the respondent does not meet the eligibility criteria assessed during screening, he or she will be routed to a page that thanks the respondent but indicates that he or she does not fit the specific criteria needed for this proposed project. The page that thanks the respondent is located in Dynata’s panel system and is outside of the actual survey.

Survey administration

Participants who meet the basic eligibility criteria will continue to the survey questionnaire (attached). The survey will ask respondents about their current levels of physical activity, what types of activities and programs are interesting to them, and additional classification questions such as highest level of education achieved, employment status, marital status, and income that will be used to segment the data during analysis. To the extent possible, these questions will be asked in the exact format as they are currently asked in the Behavioral Risk Factor Surveillance System (BRFSS).

Focus group

The data collection subcontractor, Leavitt Partners, will contract with a focus group vendor to recruit and screen focus group participants. Focus group vendors typically recruit participants from lists of individuals that they procure. The screening criteria is provided by CDC.

Eligibility criteria for the focus groups are as follows:

• Male, and

• Native American, Hispanic, or non-Hispanic Black, and

• Aged 18+

• Disease status

  • Diagnosed with diabetes or

  • Diagnosed with prediabetes, or

  • Diagnosed with hypertension, or

  • Has a family history of type 2 diabetes (parent or sibling), or

  • Is physically inactive, or

  • Aged 45 or older, or

  • BMI > 25

• No experience in market research or advertising industry

• Participated in no more than one previous focus groups

• Not a health care worker or clinician

Leavitt Partners will also procure professional focus group moderators to moderate the focus groups using a focus group guide (attached) developed by the project team. The researchers will make an effort to use a moderator from the same race/ethnicity group in order to solicit the most honest responses to focus group questions.

The guiding questions for the focus group have been developed to uncover the reasons why minority men participate in the National Diabetes Prevention Program lifestyle change program disproportionately less than women and Caucasian men. Guiding questions will focus on identifying/confirming personal challenges, vetting the promising principles identified in early work for this project, determining sources of information and support for minority men and discovering what program components are attractive to this population.

3. Methods to maximize response rates and deal with no response

Dynata uses several approaches to maximize response rates from their panelists. To prevent panelist burn-out and subsequent resignation from the panel some sources that make up Dynata’s sampling stream allow respondents to “rest”, meaning that they are not sent invitations to complete surveys they may qualify for. Some restrictions are “hard” limits and others are “soft” guidelines. There are two reasons why a panel may want to consider letting a respondent “rest”. These include (1) to avoid over-burdening them, either in terms of work done (surveys completed) or over-solicitation (i.e., filling their inbox with invitations), and (2) to prevent condition effects. Conditioning is defined as a change in the panelist, in either attitude, knowledge, or behavior, that is solely caused by their participation in market research. Dynata has replaced the normal survey invitation with a more general invitation to “give one’s opinion”. This allows invitations to be sent at a regular and consistent interval and subsequently reduce the number of emails sent. To prevent conditioning effects, respondents are excluded from a survey based on their past participation survey history. Typically, this is because they have done a survey in the same category before.

In addition to the above, Dynata employs a routing environment to efficiently allocate willing respondents to the surveys that they are best suited for and are more likely able to complete. This reduces self-selection bias associated with invitation-based methods and increases respondent satisfaction with the process. Finally, to encourage participation and thus maximize response rate, the participants who complete the questionnaire will receive a point-based incentive, valued between $2 and $4, which can be redeemed for other items.

Use of an opt-in online panel precludes us from providing an estimated response rate. The American Association for Public Opinion Research (AAPOR) believes that the best that can be provided for a non-probability opt-in panel is a “participation rate” since numbers of contacts at recruitment are unknown. One could argue that the participation rate can be measured from the point at which the respondent is formally asked to complete the study (which would be at the introduction to the survey proper). Since all respondents at this point have already made themselves available to surveys in DYNATA Dynamix, participation rates are expected to be high.

4. Test of procedures or methods to be undertaken

The survey instrument has undergone cognitive testing with 5 individuals in the respondent population. Respondents were recruited through Craigslist and Dynata and directed to a short screening questionnaire to determine eligibility. Those individuals that qualified were contacted to set up a 45-minute interview. During these sessions, study personnel allowed the respondent to work through the questionnaire. Upon completion of the full questionnaire, study personnel went through each question to determine if the question was clear to the respondent and asked them to explain what he thought the question was asking. The researchers used this information to adjust the questionnaire to improve clarity and completeness. Additionally, where applicable, we have used the same questions as they are asked on the BRFSS and NHANES survey in order to leverage the validation testing that has been completed by the BRFSS and NHANES teams.

5. Individuals consulted on statistical aspects and individuals collecting and/or analyzing data

Individuals Consulted on Statistical Aspects and Individuals Collecting and/or Analyzing Data
Name	Organization	Contact Information
Erik Krisle, MPP	Leavitt Partners	Email: [email protected] Phone: 801-326-3584
Chris Wilks, PhD MPA	Leavitt Partners	Email: [email protected] Phone: 917-538-0765
Kerstin Edwards, MPA	Leavitt Partners	Email: [email protected] Phone: 801-326-3576
Carlie Rhiness, MPH	Leavitt Partners	Email: [email protected] Phone: 517-420-7621