Every few months, Biomedical Instrumentation & Technology releases a new journal publication on recent events, changes, and perspectives happening within the Medical Human Factors community. The July/August 2017 issue features a compelling piece regarding the important role that end user and patient input has within the design process for medical devices. (Special thanks to Pat Patterson at Agilis — a partnering Human Factors consultancy here in Arizona — for the insights provided during her interview sections in the BI&T article).
If you happened to miss it, the present blog article will walk you through key information and take home messages addressed in the latest installment.
“Fail early” — (says the pot to the kettle)
The article’s author, Chris Hayhurst, sets the tone early in the article with the following statement: “It’s common wisdom in the world of innovation and design that it’s better to fail early, when mistakes are more likely to be fixed, rather than at the eleventh hour.”
This piece of wisdom is something that designers, engineers, and other professionals have heard for decades. In fact, it’s a topic that’s been addressed in our one-on-one interview with Russ Branaghan, PhD. But isn’t this mantra old news at this point? I mean, in the present state of business competition, it seems like every smart company — big and small — is arming themselves with knowledge from UX and Human Factors. We think: it might happen to those people, but it could never happen to my team. We’re too smart. Too progressive.
And yet, if we are being honest with ourselves, everyone single one of us has experienced the bitter sting on a project we have worked ourselves ragged on.
Part of the issue is that nobody thinks that what their team is doing is the wrong way as they are doing it. It’s hard to anticipate (and accept) failure if you’ve combed through details over and over again. You expect that those kinks are worked out by the time users have an early prototype in their hands (even if it’s held together by kite string and contact cement). You’ve spent so much effort just making the thing work. That is, to make it do the one thing that it’s supposed to deliver on.
Yet, within all that wheel spinning and hours of sleeplessness, you and your team overlooked the placement or color or haptics or shape or texture of a critical button. Perhaps that button gets accidentally pressed by a participant’s sweaty, fumbly fingers. Maybe it’s overlooked by a set of users’ eyes who are too worried about being watched through that giant one-way mirror.
Whatever the reason, your product and team’s efforts fall into the category of “told you so”. Should have failed earlier. Should have saved that time. Should have used those resources for something else. It’s always easy to look back on failures and see opportunities where you could have detected the issues earlier on.
With this in mind, I feel that the whole “fail early” advice isn’t all that instructive. A better piece of advice would be to organize your design process intentionally around testing early and often, rather than attempting to fit in tests opportunistically. The reason for this is that opportunities are hard to see ahead of time, especially as the design process deviates from your initial timelines. If you stick to firm commitments (e.g., bi-weekly, monthly), test what you have finished in that stage of the design cycle, and accept the results you uncover, your inevitable “failures” will come at fairly predictable time intervals.
In addition to the whole “fail early” advice offered in Hayhurst’s article, here are some tips on how to prepare a “safety net” in the beginning of a project, in preparation for failures:
1) From the beginning, expect that you will need to change every asset of your design in some way. If your design won’t allow for a button to be shifted over 1/4 “ of an inch to the right, there is a good chance that you need to rethink your design altogether. Don’t grow attached to an idea or style or placement of an asset.
2) Don’t build assets that are contingent upon users successfully using other assets ahead of time (i.e., avoid designs that permit cascading failures).
3) Test your prototype as soon as the glue dries. (This point is echoed by Hayhurst as well).
4) Build your prototype to be tougher than you think it needs to be. You shouldn’t have to tell users to be gentle when using your prototype. Let them use it how they want. If they break it, have another one ready to go. If you truly “know your users” like you think you do, you should have built your prototype (and study tasks) in ways that reflected the prototype’s capacity for wear and tear.
5) Initially design your product for your least competent user, not your average user. Make it dirt simple and easy to learn. I’ve never heard of a highly competent user complaining about a product that was too easy to use. Rest assured, however, you will add complexity to your product over time without even realizing it as the design cycle progresses.
6) Understand the variety of “contexts for use” that users will interact with your device in. For instance, if you’re designing an app for diabetes management, you cannot expect a person to attentively watch your 6 minute instructional video before they begin using the service. It’s 10:00pm. They are wearing sweatpants. They are too tired to brush their teeth. Even if they “watch” the video, they probably aren’t paying attention.
7) Identify a reliable, cheap, and (preferably) local source of end-user participants to test your device. This point is among the biggest reasons why companies don’t follow the “test early and often” mantra. Before you ever start up your CAD/CAM software, talk about packaging design, or distribution, you need to identify how you are going to recruit participants. If you don’t know the answer to this question by the time you have a prototype, you will feel the pain as you wait days or weeks for an outside agency to recruit participants for you.
8) Iteratively testing costs money up front — get over it. Get your team and stakeholders on board with this from the beginning because it will save you money in the long run. Lots of money, actually. This has been demonstrated time and time again across industries — there’s a reason why Human Factors and UX has grown so much in the last decade.
Setting the stage for patient advocacy at the FDA
According to Hayhurst, FDA begun to play an active role in advocating for a more patient-centered perspective as early as 2013. An early indication of this shift toward end-user focus came about when FDA’s Center for Devices and Radiological Health (CDRH) openly discussed its “Patient Preference Initiative” — a workshop aimed at bringing together healthcare providers, medical device manufacturers, stakeholders, and patients to talk about the challenges each group faced in the healthcare arena.
Particular emphasis was put on understanding where the patient stood in each junction in the design process. This workshop motivated FDA to create the Patient Engagement Advisory Committee (PEAC) within CDRH. Similar to the workshop, the goal of the PEAC was to create a platform for discussing and socializing patient-related concerns and issues, and to help manufacturers understand the importance of following a patient-centered approach throughout the development cycle.
Subsequently, CDRH published three guidance documents advocating for more patient-focused care:
- Applying Human Factors and Usability Engineering to Medical Devices
- Factors to Consider When Making Benefit-Risk Determinations in Medical Device Premarket Approvals and De Novo Classifications
- Voluntary Submission, Review in Premarket Approval Applications, Humanitarian Device Exemption Applications, and De NOvo Requests, and Inclusion in Decision Summaries and Device Labeling
In addition to these guidance documents, the FDA has also begun to ask device manufacturers to submit data regarding patient preferences. To date however, patient preferences remain an optional submission — one that is ultimately intended to help the agency determine whether the device is safe and necessary for patient care.
According to Hayhurst’s interview with Kathryn O’Callaghan — the assistant director for strategic programs at CDRH — a main initiative for the CDRH in 2017 will be to help the medical device industry make considerable progress when it comes to designing devices that meet real patient needs and get the outcomes that matter to patients the most.
Are there upper limits to the helpfulness of end user feedback?
Especially in light of CDRH’s push in 2017 for gathering patient preferences, it’s not unreasonable to wonder how helpful patient preferences are in respect to the medical device design process. One point drawn upon in Hayhurst’s article suggests that while patient preference is something that needs to be accounted for more often, it may present unanticipated challenges depending on the complexity and developmental stage of the device itself. For example, if a complex fMRI is being developed, when should a prototype be put in front of a patient? Should patient preference take precedent over technical details that he or she isn’t familiar with?
Keep in mind, we aren’t talking about end-users here. End-users would be trained and certified healthcare professionals. Hayhurst’s article isn’t suggesting that these individuals be kept from the development and design process. Instead, what we’re talking about are patients (indirect end-users).
From patient to “healthcare consumers”
Hayhurst’s article continues on to suggest that one area that patient preference will begin to play a more prominent and direct role is in home care settings. Unlike hospitals or other clinical settings, healthcare providers (i.e., physicians, nurses) are not usually considered the primary end-user. Instead, medical devices are intended to be used by caregivers, family members, or even patients themselves. Some interviewed suggested that home health settings will play a significant role in the future, allowing more patients to self-manage their conditions with minimal supervision — potentially due to the growing elderly population in the United States.
Jim Piepenbrink — deputy executive director of the AAMI foundation — suggests that one of the reasons why we will see an influx of home care devices under FDA review is that many of these devices were originally borrowed from hospital settings and repurposed for the home environment. Yet during this transition, little attention was put into determining if patients would be able to effectively use these devices without the assistance of trained healthcare providers. Because of this poor fit between patients and the device, many see a growing opportunity for new home care devices to compete in the market with existing products.
Alongside this burgeoning home care market is the fact that there is a significant change in how individuals perceive, approach, and use healthcare services. Hayhurst’s article describes this as the transition from a “patient” role to a “healthcare consumer” role.
The premise of this transition toward healthcare consumers is that individuals will increasingly view healthcare devices and technologies like they would day-to-day consumer technologies — ex. Televisions, smartphones. Similar to consumer technologies, healthcare consumers will seek out options that are flexible, customizable, and provide a personal fit (i.e., physically, socially, cognitively). This is possible in the consumer world because there is little elbow room in the market space. Competition amongst companies keeps pricing consistent (and usually low), and forces companies to use product innovation as a springboard for gaining market shares.
When you shift the lens toward the healthcare industry, it’s quickly apparent that some markets are more saturated than others. In fact, some markets have very few companies competing in them at all. As these markets swell with competing companies, however, Hayhurst’s article predicts that following through on designing products that meet consumer requirements, needs, and expectations will continue to be a primer driver of a company’s success.