You Need 15...so says Ron.
Understanding The Human Factor in Human Factors within Medical Device Design.
Our resident Human Factors expert @MorvenShearlaw recently spent an intensive and insightful few days in Cambridge with amongst others, the lead author of the 2016 FDA Guidance on Applying Human Factors in Medical Device Design. His name is Ron.
Ron and his fellow tag team presenter, Bob, provided some great insights and clarity on the FDA’s consideration of Human Factors in pre-market submissions.
The process presented, of frequent user testing in simulated user scenarios being critical to the safe development of medical devices aligns well with Fearsome’s market led approach.
Design development flows from user task analysis to real world observation formative testing which trial and test design features, culminating in a summative test to validate that a product is safe and effective to use. For the FDA, a summative test must test the product’s critical tasks with a minimum of 15 users from each of the products user groups.
Having lunch with Ron, Morven established that he was the very man who stipulated that 15 was the minimal number of user tests required in this process! How enlightening…even formal FDA regulatory guidelines originate from a normal, functioning human!
Ron frequently reiterated his belief and desire that human factors work should "not be burdensome’ to industry. Whilst clearly wanting to ensure patient safety, he didn’t want to put too many barriers in the way of the manufacturers doing essential, safety critical user testing. So 15 it is then.
A further piece of enlightenment came when discussing the definition of a User Group in relation to the summative testing of a products critical tasks - something the user has to be able to do, and do correctly for the product to be considered safe and effective.
An extremely useful way of delineating between user groups is this definition:
‘’A user group represents a group of people that have a significantly different interaction with a product’’.
In the case of a medical device for example, the clinician that uses the device to examine the patient could be considered a distinct group, as is the patient themselves.
Whilst the patient group may vary wildly, it is only when these variations affect their product interaction that they are considered a clear-cut distinct user group.
There can be a tendency (guilty!) to over specify these groups in the pursuit of thoroughness and to ensure that we’ve captured the user experience for children/ adult/ elderly, male/female, different states of patient e.g. a patient in distress/ not in distress. During a team exercise Morven and her group had to design, execute and document an example summative test for an Epinephrine Auto-injector. The group reasoned that a person going into anaphylactic shock and using it on themselves was a different user group to those administering it on a person going into shock.
Within these two distinctions they further divided the groups by age and level of training, creating 10 separate groups. With 15 participants for each group this would require 150 separate (functional) tests to be run, observed, analysed and documented and a significant budget to do so.
Thorough work as pointed out by Ron and Bob... but potentially way over the top!
Morven’s group had created more work for themselves, increasing the cost burden for their (fictitious) company by not thinking carefully enough about the process and the critical tasks involved. In the case of the autoinjector, using it yourself or using it on someone does not require a significantly different interaction with the product. Understanding this halved the tests required…and the budget to do so.
The FDA is a large organisation with significant power and responsibility and can often come across as a scary, faceless corporate entity.
This invaluable course allowed human connections to be made and a realisation/ reminder that there are actually individuals who create these guidelines, read our submissions and make decisions.
Thanks Ron…and Bob…and Morven!