This poll was conducted April 23-May 1, 2024. A sample of 1,479 U.S. adults were interviewed online and by telephone. The estimated margin of sampling error for the entire sample is plus or minus 3 percentage points. The margin of error may be larger for smaller groups within the full sample. Sampling error is only one of the potential sources of error in a survey.

This poll was conducted April 23-May 1, 2024. A sample of 1,479 U.S. adults were interviewed online (1176) and by telephone (25). The respondents are members of the SSRS Opinion Panel, recruited through a national ABS sample of U.S. adults aged 18 or older and through random digit dialing. An additional 278 telephone respondents were sampled through random digit dialing of prepaid cell phone numbers supplied by Marketing Systems Group, targeted to contact Hispanic and non-Hispanic Black respondents and others who are less likely to participate in online panels. The estimated margin of sampling error for the entire sample is plus or minus 3 percentage points. The margin of error may be larger for smaller groups within the full sample. Sampling error is only one of the potential sources of error in a survey.

Interviews were conducted in English (1396) and Spanish (83). The average time for online interviews was 15 minutes; the telephone interviews averaged 31 minutes. Monetary incentives ($15 for phone and $5 or $10 for online interviews) were offered to sampled individuals to increase participation.

These are the exact questions asked in the interview that were discussed in the news article:

• "How much have you heard, if anything, about a class of drugs being used for weight loss, such as Ozempic, Wegovy, and Mounjaro?" [A lot, Some, A little, Nothing at all]
• "Are you currently using or have you ever used one of these drugs to lose weight or treat a chronic condition such as diabetes or heart disease?" [Yes, currently using; Yes, but not currently using; No, never]
"Do you take these drugs primarily to lose weight, or to treat a chronic condition like diabetes or heart disease, or both?" [To treat a chronic condition, To lose weight, To lose weight and treat a chronic condition]
• "How did you pay for the cost of these drugs?" [Paid the full cost themselves, Insurance covered part of the cost and respondent paid the rest, Insurance covered all of the cost]
• "Did you get these drugs or a prescription for them from any of the following places?” (select all that apply) [Your primary care doctor or a specialist; A medical spa or aesthetic medical center; An online provider or website; VA doctor/clinic (volunteered); Somewhere else]

This poll was conducted April 23-May 1, 2024. A sample of 1,479 U.S. adults were interviewed online and by telephone. The estimated margin of sampling error for the entire sample is plus or minus 3 percentage points. The margin of error may be larger for smaller groups within the full sample. Sampling error is only one of the potential sources of error in a survey.

Key challenges in conducting a poll are 1) selecting a representative sample, 2) convincing sampled people to participate in the survey, 3) attending to potential errors in data collection and 4) adjusting the sample to make it more representative. This is how KFF worked on these problems in this poll:

Selecting the Sample. Respondents for the KFF Health Tracking Poll were obtained from an online panel and telephone contacts. The online respondents (1,176) and some telephone respondents (25) were members of the SSRS Opinion Panel, a sample of U.S. adults aged 18 or older. SSRS selects addresses from a national list maintained by the U.S. Post Office, giving each address a known chance of being selected. This procedure is called a “probability-based,” method. Unlike volunteer or "opt-in" panels, this sampling method allows those who use the SSRS panel to estimate the representativeness of their surveys – a "margin of sampling error."

SSRS sends invitations to people living at selected addresses to join its panel of respondents. The individuals who agree are then sampled periodically and invited to participate in surveys, such as the KFF Health Tracking Poll.

Because some households in the U.S. do not have internet access or are otherwise hard to reach, SSRS also recruits panel participants by telephone, using a sampling technique called “random digit dialing.” For the Health Tracking Poll, KFF added more telephone interviews to the SSRS panel. They obtained 278 additional interviews obtained through another probability-based method – a random digit dialing telephone sample of prepaid cell phone numbers obtained through a company called Marketing Systems Group. Phone numbers used for this prepaid cell phone component were randomly generated from a cell phone number list targeting Hispanic and non-Hispanic Black respondents and others who are less likely to participate in online panels.

Urging People to Participate: One of the other potential sources of error is "nonresponse bias." Some of the people who are invited to participate in a poll do not agree to do so. If those who do not agree to participate are different in pertinent ways from those who do agree, then the survey findings may be biased. Researchers try to persuade invited people to respond to reduce the likelihood of bias. In the KFF Tracking Poll, invitations to respondents included monetary incentives ($15 for phone respondents and $5 or $10 for online respondents) to increase participation. In addition, multiple attempts were made to contact and encourage both online and telephone respondents to participate.

Addressing Data Quality: The questions used to measure respondents' opinions and experiences, and how they are administered, can introduce error into the survey. KFF employed several methods to reduce the chance of errors. First, the questions were tested prior to the poll to see if they were understood. To introduce the questions in the survey, examples of drugs used for weight loss were provided so that respondents would be sure to know what the questions referred to. These are the exact questions asked in the interview and discussed in the news article:

• "How much have you heard, if anything, about a class of drugs being used for weight loss, such as Ozempic, Wegovy, and Mounjaro?" [A lot, Some, A little, Nothing at all]
• "Are you currently using, or have you ever used one of these drugs to lose weight or treat a chronic condition such as diabetes or heart disease?” [Yes, currently using; Yes, but not currently using; No, never]
• "Do you take these drugs primarily to lose weight, or to treat a chronic condition like diabetes or heart disease, or both?" [To treat a chronic condition, To lose weight, To lose weight and treat a chronic condition]
• "How did you pay for the cost of these drugs?" [Paid the full cost themselves, Insurance covered part of the cost and respondent paid the rest, Insurance covered all of the cost]
• "Did you get these drugs or a prescription for them from any of the following places?" (select all that apply) [Your primary care doctor or a specialist; A medical spa or aesthetic medical center; An online provider or website; VA doctor/clinic (volunteered); Somewhere else]