Dear FDA

Dear-FDA-Panel

by Eric F Shaver, PhD

Earlier this month, the Human Factors and Ergonomics Society (HFES) held the 2016 International Symposium on Human Factors and Ergonomics in Health Care: Shaping the Future in San Diego, California.  The event was well attended by researchers, practitioners, industry representatives, and students from around the world with an interest in advancing human factors efforts in the healthcare domain.  

On the closing day, Tim Reeves, PhD CHFP, Managing Director of Human Factors MD, moderated the Dear FDA panel. The idea for the panel was born out of an impromptu discussion at last years symposium. As Tim outlined at the start of the session, five years have past since the FDA published the Draft (and now Final) Guidance. And while the Guidance has been tremendously helpful in improving the safety and effectiveness of medical devices, there are opportunities to enhance the FDA’s approach: to improve the quality and validity of the human factors data we collect and conclusions we draw from human factors studies. Five years of practice is also a sufficiently long time to begin to ask the broader question of whether the FDA should redefine the boundaries of its approach to human factors?

In formulating the panel, Tim challenged panelists with the task of making a case for changing some aspect of the FDA’s current thinking on human factors. The arguments were to be science-based: panelists were asked to provide a solid rationale, based on human factors principles, for their recommendations.

The panel consisted of four human factors experts:

  • Adam Shames, MBA – Founder and CEO of Core HF
  • Eric Bergman, PhD – Director of Human Factors Engineering at Fresenius Medical
  • Shannon Clark, BS – Principal of UserWise
  • Anthony (Tony) Andre, PhD CPE – Owner and Founding Principal of Interface Analysis Associates

Below are key points covered by each panelists:

Adam Shames, MBA

Adam argued that the FDA should expand its focus to incorporate other aspects of user-centered research and design, beyond safety.  While patient and user safety is a necessary objective, other aspects of user experience have the potential to influence user behavior in meaningful ways, including the rate of device adoption and compliance with medical therapies.  Adam argued that broadening the FDA’s focus, beyond safety, would encourage sponsors to develop products that are more apt to meet all patient and user needs.

Eric Bergman, PhD

Eric recommended that the FDA reconsider its definition of “use error” offered in the final Human Factors Guidance. The definition limits “use errors” to use issues with implications for safety, and as such, is inconsistent with the broader definition of use error provided in 62366-1. Eric also asked that the Agency clarify how sponsors should think about “effectiveness” when attempting to validate “safe and effective” use.  Would the Agency approve a device that was safe but not effective?

Shannon Clark, BS

Shannon recommended that the FDA allow sponsors to consider the likelihood of use-related hazards when defining critical tasks and risks. While the Guidance directs sponsors to focus on criticality, as likelihoods are inherently difficult to assess, Shannon made the point that the very concept of “reasonably foreseeable” hazards requires consideration of likelihoods, and accordingly, consideration of  likelihoods is inescapable.  We still need likelihood to determine test cases for validation.

Tony Andre, PhD CPE

Tony raised two issues relating to human factors validation testing. The first is that some products are best validated under more longitudinal use conditions, or after an initial learning phase, rather than during first exposure. The FDA’s focus on initial use may uncover initial challenges when learning to use a device, but doesn’t validate that devices can be used safely once learned.  The second questioned the value in testing untrained users in validation studies without asking them to read the instructions. Tony argued that this use case doesn’t tell us much. How can a study of what happens when people don’t get trained and don’t read instructions be a validation?

The event was attended by approximately fifty people, and included a robust exchange between the panelists and members of audience. Among those in attendance were Irene Chan, the Deputy Director of CDER’s Division of Medication Error Prevention and Analysis (DMEPA), and Xin Feng, from CDRH’s Human Factors Pre-Market Evaluation Team. Both Irene and Xin thanked the panel, welcomed the discussion and offered comments and insights of their own.