March 2013

Summary

From August 2012 through December 2012, RTI International executed a Web-based survey and conducted in-person interviews with a total of 590 health care practitioners, including prescribers, RNs and LPNs, and technicians and therapists living in the United States. RTI International randomly assigned Web-survey participants (N = 574) to view and provide comments on one of three sample versions of abbreviated device labeling developed for a hypothetical fluid infusion pump. In-person interviewees (N = 16) answered the same questions as the Web survey participants along with additional probes to elicit more in-depth responses. Labeling versions reflected the design suggestions from formative research findings and differed only in format and section ordering.

Conclusions and Recommendations

The survey of health care practitioners offered a number of insights regarding perceptions of medical device labeling both in general and for the example prototypes presented to survey respondents. One important highlight from this study is the highly positive sentiment expressed by most respondents regarding the labeling prototype presented. Many respondents reported they would be likely to use such labeling in a variety of circumstances and reported overall satisfaction with the presented labeling. In-depth, in-person interviews with a small number of practitioners also confirmed most of the larger patterns suggested by the Web survey.

The report compared responses from prescribers, RNs and LPNs, and technicians and therapists. The report also compared responses from those with relatively more and less professional experience. One prominent theme here is that there are relatively few differences between groups in their responses to labeling. There were some noteworthy differences; more experienced participants were somewhat more likely to find the labeling to be generally useful, for example, and RNs and LPNs expressed slightly more confidence in using the device after reading the labeling than did respondents in the other job category groups. In general, the various respondent groups tended to view the labeling in a similar fashion.

All three versions of the labeling received a largely positive response in terms of perceived general utility and in terms of the confidence afforded by using the labeling. Moreover, respondents also tended to view the versions of labeling prototype as equally useful.

Results regarding format preferences also yielded a general sense of satisfaction with the basic structure and order of the labeling prototype. More than 90 percent of the respondents (524/574) were satisfied with the manner in which information was ordered on the prototype. Respondents did have some specific recommendations as well. For example, respondents preferred bulleted warnings and precautions and preferred numbered instructions and most respondents recommended always placing words such as “not” and “only” in bold type. Overall, participants also tended to recommend that a table of contents be included for labeling that is five or more pages long, although those who saw a table of contents in their version of the labeling prototype tended to say that a table of contents should be included all the time.

In addition to format, participants largely approved of the labeling prototype content as well, although a few respondents noted possibilities for minor improvements such as including additional safety information related to using the device in specific circumstances (e.g., with special populations, near flammable gases, in the presence of an MRI, or with rechargeable batteries). Useful next steps for prototype development could include addition of practical information describing the required battery for a device as well as information on what to do with the device when it is not in use and how to ensure that the device continues to function over a period of time.

Virtually all users thought that having labeling appear with or on the device is a high priority. We also learned additional information about respondents’ preferences for labeling medium. For many users, such as nurses who regularly use devices, access to hard copy labeling that appears with or on the device is a higher priority than supplemental labeling that might appear elsewhere. Beyond conventional hard copy labeling on or with the device itself, there are additional opportunities to develop means for accessing labeling because a substantial number of respondents expressed desire for electronic labeling or labeling available via a Web site. We also noted what may be a generational shift in preferences. Practitioners with less than 10 years of experience, for example, were less likely than their more experienced counterparts to opt for labeling that is physically appended to the device and were more likely to call for labeling to appear via a Web site.

Both a Web-based survey of health care practitioners and in-depth, in-person interviews suggested the utility of the labeling prototype developed by project staff. Key material appears to be presented in a satisfying order, and respondents appeared to feel reasonably confident about their use of the device after reading the labeling. Such labeling, whether available in hard copy with the device or through electronic means, holds promise to bolster device user comprehension and confidence.