FOCUS GROUPS (continued)
Panelists in both cities and both age groups largely described themselves as competent, seasoned and careful drivers. They saw the experience of driving as being more difficult today than in the past, but assigned blame for that to the growing carelessness (e.g., cell phone users) and the aggressiveness (e.g. speeders, tailgaters) of other drivers, rather than to their own diminished skills.
Only after considerable probing were the panelists willing to admit any loss of skill, except for problems with night vision. Only panelists over age 75 were willing to admit any real limitations; however, even they were more likely to describe themselves as more patient and cautious than in the past, rather than more fearful or concerned about their own response time.
Among the more interesting and useful products of focus group discussions are the actual words of the panelists. Included in this summary are sample comments from the eight hours of taped discussion that best represent the attitudes that emerged from the groups. The quotes are coded so they can be referenced to a specific group location and age cohort. The first letter of the code, either a “T” or a “P,” indicates Tampa or Philadelphia. The second letter of the code, either a "Y" or an “O,” indicates the younger group (55-69) or the older group (70+).
The moderator explained the purpose of the proposed studies and the general methodological parameters to be explored. By direct and indirect questioning and by careful listening, an assessment was made regarding the willingness of the panelists to reveal to a stranger their personal array of medications.
Panelists indicated almost total willingness to share information about their medications. Although never queried about it directly, individuals volunteered a large amount of information about the drugs they were taking in the course of the discussions. This was true in both cities and both age groups. However, throughout the process of sorting methods into two groups (acceptable/would participate vs. unacceptable/would not participate) fewer older drivers perceived as many of the methods as acceptable as the younger drivers. This means that it will likely be much harder to recruit participants age 70 or older in future research on medications and driving than those in the 55-to-69 age group.
With some probing, two overall concerns emerged about the prospective research—one that might be a real barrier to participation, the other a factor that might reduce interest in it.
Willingness to Participate
This discussion section began with a review of the goals and objectives of prospective future studies, and by reiterating assurances of confidentiality and no impact on driving privileges. Then, each of the key methodological variables was introduced and discussed, one by one. Finally, a paper study was conducted in which panelists could select and rank-order their preferences among the methodologies they found acceptable.
Methods 1 through 3. Bringing all medications in a bag to (family doctor, pharmacist or researcher) to (an office, pharmacy or research facility) so that a complete list can be made and a discussion can take place about when and how medications are taken.
The medications-in-a-bag approach was generally well received by all groups and should be considered a strong candidate for conducting future research. This was one of the two highest-scoring methods. However, it appears that the person conducting the medication reviews is likely to have a major impact on the degree of participation.
Many panelists expressed a slight preference for simply preparing/bringing in a list of their medications, dosages, etc. The reason for this is that at least half the panelists already carry a list of their prescription medications on them at all times. In the older groups, as many as 70 percent had such a list on their persons. These people preferred to use the list. Some suggested that those being surveyed be given an option (i.e., bring their medications in a bag or bring a list of medications).
Regarding the type of person who should conduct the interview, all three choices were acceptable to most panelists. However, there were some strong, clear preferences. The most acceptable choice, by a wide margin, was a pharmacist, followed in order by a doctor and a professional research person. Several reasons were cited.
Again, panelists did not rule out the other choices, they were only less comfortable. Doctors have some obvious advantages from a qualifications and empathy standpoint. Researchers, though, were seen as more neutral, less likely to judge or threaten in some way.
Likely-to-participate percentages also varied by age group. Pharmacists were deemed acceptable for this role by the highest percentage of participants overall at 63 percent (70% among younger; 55% among older). Doctors were deemed acceptable by the next highest percentage of participants—by 55 percent across age groups (62% among younger; 47% among older). Professional researchers were the least acceptable personnel among the three types, with 46 percent of participants overall indicating that they would be willing to participate (55% among younger; 42% among older.)
Methods 4 through 7. The technical options are exactly as described in Methods 1 through 3 except that the interview would not be conducted in an office, pharmacy, etc., but in the person's home. Also, the option was added to have the interview conducted by a nurse or an occupational therapist (OT).
Reactions to this series of options were strong, spontaneous and highly negative. Two-thirds were strongly opposed to it. It appears that every concern and suspicion that panelists might have about the study was instantly brought forth in response to holding an interview in their homes—confidentiality, breach of privacy, personal security, etc.
While it appears unlikely that doing the research in subjects’ homes would draw much participation, the question of who the panelists trusted more to come into their homes was still discussed. Generally, the most trusted (or least threatening) figure was the nurse. Overall, 46 percent approved of a nurse conducting research in the subject's home (35% among younger panelists and 57% among older panelists). Next highest in approval was the pharmacist, by 38 percent overall (27% among younger panelists and 41% among older panelists). Third in acceptability was the professional researcher, with 24 percent approving this method overall (12% among younger panelists and 35% among the older panelists). Last was the occupational therapist, with 12 percent overall approval—identical for both younger and older groups.
Method 8. Fill out a survey that the researchers mailed to you, asking you to list all the medications you take, the dosages and how often you take them.
Even though this method was perceived as the easiest, only about half the panelists said they would cooperate with this approach. The major drawback seems to be a fear that survey results would fall into the wrong hands, because the method is too anonymous to be trusted.
The moderator then suggested that participants be given a code number for the questionnaire, instead of signing the questionnaire (to preserve confidentiality). This improved the rate of promised response. But, some panelists still doubted.
Younger drivers are more likely to participate in a written survey (57% said “yes”) than older ones (35% said “yes”).
Methods 9 and 10. Come to the researcher's office (alone or with a companion) without bringing the medications with you. The interviewer would ask you to remember your medications and dosages.
Method 9 was very popular, receiving a positive vote by 65 percent overall, the highest of any listed method. Second highest was bringing in the bag of medicines to a pharmacist (acceptable for 63% overall). In general, panelists liked the ease of it. Younger panelists were more likely to support this method (75%) than older panelists (55%). Older panelists were more worried that they would not be able to recall all their medications. Allowing the older panelists to bring in a list resolves their problem.
Bringing a companion (Method 10) introduces issues of privacy and drops the rate of acceptance. Fifty-two percent of younger panelists say they might do this while only 37 percent of older panelists would.
Method 11. Give your consent to a researcher to send a letter to all of your doctors, asking them to list all the medications they prescribed for you. The doctors would send that list back to the researcher.
Based on these focus groups, there isn't the slightest chance of getting significant cooperation from drivers if this method were employed in the research. It received, by far, the lowest willing-to-participate score (19%) and was equally disliked in both cities and age groups.
Method 12. The researcher obtains a list of medications that have been prescribed to you. You then bring in your over-the-counter medications to add to the list. The researcher then talks to you about all the medications.
The response to this method is likely an aberration, or misunderstanding of the intent of the method. After vehemently objecting to the idea of obtaining the list of medications from a third party, panelists here say that they would participate (51% overall, 57% older panelists and 45% younger panelists). This response should probably be interpreted to mean simply that they are willing to bring in their over-the-counter medications.
Method 13. Having special caps placed on the medications you take that record the date and time each time you open the container.
Panelists genuinely disliked this idea. They found it inconvenient, impractical and annoying. Seventy-one percent said that they would not participate in a study in which this were the central methodology.
Method 14. Researchers would obtain information about the prescription drugs you take by accessing pharmacy databases.
Only 34 percent of all panelists would agree to participate in a study using Method 14, with more older panelists likely to participate (40% say “yes”) than the younger ones (27% say “yes”). The existence of the pharmacy databases had not occurred to most of them before, and they found the fact disturbing.
Each panelist was provided with a stack of cards, each containing one of the fourteen methods discussed above. They were instructed to divide all the cards into two piles: one representing “I would participate in the study,” and the other representing “I would not participate in the study.” Then, they were to draw an “X” across all cards in the negative pile. The positive cards were then to be put in preferential order and numbered starting with a “1” to indicate first choice. Table 1 presents the percentage of drivers willing to participate in research to identify medication use, for each of the 14 methods, as a function of focus group location and participant age group. The mean rank is provided for each method. Figure 1 provides the willingness-to-participate percentages by age group and overall, for each method, and Figure 2 provides the mean rankings of preferences, among the methods deemed acceptable.
Once again, the moderator provided the panelists with a broad-brush idea of the prospective research, focusing attention on the need to observe and measure driving performance after taking their usual (combination of) medications. All groups indicated a general willingness to participate in the study as outlined.
This time, though, the level of enthusiasm was less and the level of concern greater than in the previous phase. Generally, panelists perceived a greater opportunity to “look bad” and perhaps even to be penalized for poor performance. Some worried about safety during the study. Many questioned the ability to control all the variables in the study and to produce a reliable finding.
Based on these groups, it appears that people would participate, but with a greater degree of pre-sell needed and, possibly, greater incentives. The key factors emerging from the discussions that would encourage participation are:
Willingness to Participate
This discussion section began with a review of the goals and objectives of the research, and reiteration of assurances of confidentiality and no impact on driving privileges. Then, each of the key methodological variables was introduced and discussed, one by one. Finally, a paper study was conducted in which panelists selected and rank-ordered their preferences for methods that they found acceptable.
Methods 1 and 2. Method 1: You drive a vehicle with dual controls on a closed course and either a driving instructor or an occupational therapist sits in the passenger seat while you drive the car. Method 2: Same technique, but test is conducted in traffic, not on a course.
Panelists were comfortable with the idea of testing in the dual-control car, depending on where the test was conducted. Testing on a driving course generated willingness to participate by 59 percent of the participants. Testing in traffic dropped the percentage of discussants who would be willing to participant to 31 percent—the lowest of any method tested.
The striking thing about the willingness-to-participate percentages is that the older group was more comfortable driving a test car with dual controls in “live” traffic (50% positive) than the younger group (12% positive).
Methods 3 and 4. Method 3: Drive your own car on a closed course with a driving instructor or occupational therapist in the passenger seat. Method 4: Same technique, but test is conducted in traffic, not on a course.
The concept of driving their own car (as opposed to a dual-control car) was viewed more positively overall, primarily reflecting the Tampa group’s pattern of response. There was little difference by age. Sixty-three percent of the panelists indicated that they would participate if Method 3 were employed in the research (57% of the Tampa respondents and 67% of the Philadelphia respondents). Looking at the acceptability of Method 1 by location, only 22 percent of the Tampa respondents would participate if a dual-control car were used, compared to 95 percent of the Philadelphia respondents.
Even in Method 4, involving actual traffic, 54 percent across locations and age said “yes” to participation, although still more Tampa respondents than Philadelphia respondents were comfortable with driving their own cars than a dual-control test vehicle. Clearly, the comfort level of driving their own car is important to many.
Method 5. You drive your own car in traffic and miniature audio and video recording instruments are mounted in the vehicle to record the driving scene and how you respond to it.
The response to this idea was controversial. Responses ran from out-right opposition to genuine enthusiasm. Many “Big Brother” comments were made. On the other hand, some found the idea even less intrusive than the other observation methods, because another person is not in the car and because one is likely to forget the presence of the recording devices over time.
The Tampa groups were much more enthusiastic than the Philadelphia groups—65 percent of the Tampa group and 35 percent of the Philadelphia group would be willing to participate. Overall, though, panelists were split 50%-50% with regard to the acceptability of Method 5.
Method 6. A study method to briefly measure your vision, memory, and physical abilities needed to drive safely, using either paper-and-pencil tests, or tests that might be presented on a computer.
This was the most popular method tested, but still only one of several that might be appropriate to implement in future studies. Overall, 69 percent of panelists indicated that they would be willing to participate in research using this method. Seventy-five percent of the younger group was positive about this method, compared to 60 percent of the older group. These are high percentages for both, but the higher percentage of younger drivers indicating willingness may be explained by greater comfort with computers among this age group.
Each panelist was provided with a stack of cards, each containing one of the six methods discussed above. They were instructed to divide all the cards into two piles: one representing “I would participate in the study,” and the other representing “I would not participate in the study.” Then, they were to draw an “X” across all cards in the negative pile. The positive cards were then to be put in preferential order and numbered, starting with a “1” to indicate first choice.
Table 2 presents the percentage of drivers willing to participate in research to determine the effects of their medications on driving performance, for each of the 6 methods, as a function of focus group location and participant age group. The mean rank is provided for each method. Figure 3 provides the willingness-to-participate percentages by age group and overall, for each method, and Figure 4 provides the mean rankings of panelists’ preferences among the methods that were deemed acceptable.
The biggest concerns were loss of driving privileges as a consequence of poor performance on the driving assessment, and a fear that confidential information could fall into the wrong hands. A few participants commented that they know their medications can make them dizzy, and if they felt dizzy (or otherwise impaired), they probably would not participate in the driving evaluation phase. This raises a question about a bias that could result if those people whose driving is most affected by (multiple) medication use select themselves out of the study.
Participants would expect to be compensated. Nearly all panelists expressed a preference for cash, although a willingness to accept other kinds of remuneration, especially free meal coupons, was indicated.
Participants commented that the demands on their time in the medication identification phase would be significantly less than in the driving phase, and that study compensation would need to reflect that.