| Comment Record|
Ms. Michele Hardy ||
2001-05-21 09:15:48 |
| Comments for FDA General |
5. General Comments
May 18, 2001
Dockets Management Branch (HFA-305)
Food and Drug Administration
5630 Fishers Lane
Rockville, MD 20852
Re: Docket No. 01N-0078
Assessment of Physician and Patient Attitudes Toward
Direct-to-Consumer (DTC) Promotion of Prescription Drugs
GlaxoSmithKline appreciates the opportunity to comment on the proposed information collection that follows the agency’s 1999 patient survey and expands the information collection to include physicians. Our understanding of the purpose of the surveys is to obtain information from physicians and patients about their reactions to, and behaviors that stem from, DTC prescription drug advertising in order to develop policy on appropriate requirements for regulating drug product promotional materials. Given these objectives, it is critical that the surveys are conducted in such a manner as not to introduce bias or influence the outcome. To this end, we have the following comments on the surveys.
The Federal Register Notice references a prenotification letter that will be mailed to all potential respondents. The content of this letter is important so as not to bias the sample of physicians that participate in the survey. Because a copy of this letter was not made available, we are unable to comment on the appropriateness of the letter.
Are there any subquotas by specialty for the physician arm of the research?
Are there any subquotas by geography, gender, age, race, income level, etc. for the consumer arm of the research?
How will the survey data be weighted so that it accurately reflects the universe of physicians and universe of patients?
What are the expected confidence bands around the parameter estimates generated from this survey?
Telephone Interviewing: In general, research conducted with an interviewer will yield more positive responses or agreement to the questions asked than for the same research conducted utilizing a self-administered questionnaire. For example, a higher proportion of respondents would “agree” with an “agree/disagree” question if that question were asked in a personal interview rather than a telephone interview. Similarly, a higher proportion of respondents would “agree” with an “agree/disagree” question if that question were asked in a telephone interview rather than a self-administered questionnaire. This should be kept in mind when interpreting the results of this research. We suggest that it would be more appropriate to compare the relative results among different questions in the survey than to look at the absolute results from any one question.
Two Questionnaire Versions: We suggest two different versions of the questionnaires for questions 33-43 in the consumer survey, and questions 23-30 and 39-50 in the physician survey. The second version would contain questions currently worded in a positive sense, reworded in a negative sense, and questions currently worded negatively should be reworded positively. For example, Question 23 in the physician study currently reads, “I felt pressured to prescribe a drug for this patient.” We suggest a second version of the questionnaire where the question would be reworded, “I did not feel pressured to prescribe a drug for this patient.” The results from the two versions of the question should be compared to see whether or not they are consistent. If the proportions “agreeing” from one version of the question are the same as those “disagreeing” to the other version of the question (and vice-versa) then the results from this question constitute a true finding. If, on the other hand, the proportions “agreeing” or “disagreeing” from the two versions of the question are the same, then we should conclude that the question is biased and leading in either form, and the results from the question are therefore suspect.
Open-ended Questions: We strongly recommend making questions 19, 27 and 32 in the consumer study into unaided, open-ended questions. The responses listed currently are leading and biased. Changing these to open-ended questions will ensure that the information obtained would be evoked from the respondent on his/her own, rather than suggested by the survey instrument itself. This change should also be considered for
question 4 in the physician survey for the same reasons previously discussed. If physician question 4 ultimately remains as it is currently worded, a response for “other” should be added.
Other Recommended Question Modifications:
Section III. Interaction with Doctor This entire section must appear in the survey prior to Section II Attitudes Toward Prescription and Over-the-Counter Drugs. Section III elicits various reasons that patients make appointments to see physicians, one particular reason of interest being DTC advertising. Since Section II asks about a dozen questions related to DTC advertising, placing that section in the questionnaire prior to Section III will undoubtedly bias the results obtained in Section III. This change in the order of the questions is vital to obtaining unbiased, accurate results about consumers’ impressions of DTC advertising.
Questions 21 – 23, 32: These questions are potentially problematic. Instead of asking the respondent for his or her opinion about an issue, the respondent is asked to speculate on what someone else’s opinion might be or would have been. In addition, questions 21 and 23 provide leading, ready-made answers for the respondent that would lead to biased results.
If Question 21 is retained in its current form, the list of possible responses appears to be incomplete. Responses should be added for “Stop your current drug”, “Add another drug to the drug(s) you are currently taking”, and “Change the dose of your current drug”.
Question 26: This should be split into two separate questions. It is confounding as it is currently worded, asking whether they mentioned an ad, and whether they brought any information along with them.
Question 42: Should the word “are” be changed to the word “should”? Is the goal to measure awareness of current governmental policy or to understand what people think that policy ought to be?
Question 6: We suggest rewording the question “…or remained the same over the last year.” We think that two years is too long a time period for physicians to remember accurately.
Questions 39-45: We would suggest adding several additional questions:
The patient would/would not have otherwise sought medical treatment for his or her condition.
Overall, how would you say DTC advertising has affected the quality of patient care? The same response scale in Question 45 should be adopted for this additional question.
These comments reflect GlaxoSmithKline’s desire to help the agency field studies that appropriately address FDA’s need for information on DTC prescription drug advertising. Because this information is going to be used to develop policy for regulating drug product promotional materials, it is crucial that the data are collected in a rigorous manner. We believe our comments will enhance the quality of the information that is being collected. We want to thank the agency for taking these comments into consideration as these questionnaires are finalized.
Michele M. Hardy
Head, Strategic Product Labeling