Deploying an Agile Method in Digital Medical Device Design

A patient-centric approach to designing digital medical devices delivers clinical impact and drives value by empowering the patient to participate in their own care, according to Jennifer Miller, MD, PhD, physician consultant, McKinsey & Co. Miller spoke in the MD&M BIOMEDigital panel discussion, “Design & Product Development Principles for Connected Devices.” During the session, industry experts discussed all aspects of agile product development and stressed the benefits of involving and considering the needs and limitations of all users into the design process of medical devices.

In this agile method of product design, Miller said the first thing to do is to frame your problem and design, and identify your user. “You then develop out your user journeys to really understand in an end-to-end manner what the product that you are developing is going to be doing,” she explained. The next step, she said, is assembling an agile team to design the initial concept to prototype and test with select customers, incorporating their feedback to develop a minimally viable product, which is then introduced to a wider set of select customers to incorporate their feedback to develop a product program. “Finally, you scale this to the live customers,” she said.

Since defects in medical devices can lead to death, safety is a top concern, she noted. Companies need to create a quality management system early on and throughout the product life cycle, Miller said. Other things to consider are that regulatory compliance should be appropriate based on the particular agencies that will dictate a product, and to identify possible cyber security risks and build in data privacy protections.

Panelist Alisa Niksch, MD, pediatric electrophysiologist and medical device executive, Tufts Medical Center, chimed in on the topic of design feedback and user adoption, sharing tips on how to practically use a design strategy to engage patients and clinicians in the device-building process. Like Miller, Niksch said the first step is choosing a problem to solve, or a population to impact, which will guide regulatory decisions, including what pathways the device fits into.

“There are quite a few predicates out there, which is good news for a lot of designers in terms of medical devices and biosensors,” she commented, “so definitely utilize them as resources of how they conducted their studies, and their feasibility trials as well.”

She stressed that clinical trials are an extremely important part of getting your product adopted by clinicians. “Physicians are a tough audience; nurses are a tough audience,” she said. “We're very interested in regulatory pathways; we are very interested in your data accuracy. We're also very interested in integration into our electronic health records and HL7 standards,” she explained.

Design goals for patients should also include disability access and physical comfort to make the device versatile and workable in these populations. To get more information about what the needs are, she suggested spending time with as many patients as possible, such as at a medical center where you can also gain insights from physicians, and also with family members who might be using your product at home.

Amy Bucher, PhD, vice president, behavior change design, Mad*Pow, shared her perspective on why behavior scientists really belong in the digital health development process. “The key thing to think about is what makes [a product’s] effectiveness meaningful,” she said, noting that outcomes, the results of using a digital health product, are the product of both effectiveness and engagement.

Bucher covered some frameworks in behavioral models that might be used in behavioral design and experience strategy, pointing out that they all have an engagement component. “How do people become attracted to the idea of performing a behavior? And what do they need to have present in their situation in order to sustain that behavior?” she asked. “So as we look at developing digital health products, I would argue that we need to similarly take that sort of lens and look for opportunities to let people exercise these psychological needs.”

Taking a holistic approach and bringing all these perspectives together, Jaydeo Kinikar, vice president, virtual care offering management, Best Buy Health, said: “It is really important to ensure that you are thinking holistically about customer needs—clinical, financial, operational,” Kinikar said. “And not only about a specific customer audience, think about different clinicians, think about the end-user, a patient, and a caregiver.”

When Miller asked how one would create a device or platform that is both safe and agile, Niksch answered that this is where partnerships with hospitals and medical groups and large healthcare systems can help, especially with telehealth technology. “And we've seen that already supported by the federal government with some FEC grants that went out to places like the Mayo Clinic, who managed COVID patients at home,” she said.

“You really want to ensure that you're clear about [the] type of medical device,” Kinikar said. Depending on the use case and the type of data you are collecting as well as the insight you are driving, he said, a device might not need to be a completely class 2 device.

The second important thing is to test a device early on in a clinical setting, Kinikar said.

“When you're just thinking about that engagement layer, about the coaching, the support, the behavior change that goes alongside these sorts of devices, there's a great advantage to being agile. One of the things that we know is that human behavior is incredibly context driven, and as context changes, having an agile software development method in play allows us to be responsive,” Bucher said. “One of the best examples that I think we're all familiar with now is with the COVID pandemic, where we suddenly saw a very dramatic shift in the way that people were using telemedicine.”

Mentioning the recent implementation of FDA’s Center for Digital Health, Miller asked the panel what the shifting regulatory scene may mean, and if this might be an opening for young startups.

“I have noticed that smaller companies are often reluctant to venture into areas that are FDA regulated because there's a perception that it's very complicated,” said Bucher. However, she said that the understanding of digital health has become more sophisticated and the process is becoming a little bit clearer.

“I think we're going to see a lot more and more of ways of working with the FDA,” Kinikar said. “The pandemic has been a great disruptive force to really open the doors, not only at FDA, but at different providers and payers and different industry players, because the barriers that we used to perceive as unsurmountable, I think are coming down.”

Niksch commented that FDA is fairly sophisticated with AI. “They understand the limitations of it, but they are treating it like any other technology, and they want data and they want an actual verification process, much like any other piece of software or medical device,” she said.

Finishing up the discussion, Miller asked about challenges and successful strategies the panelists have used to incorporate this user design approach.

Kinikar stressed that it is critical to incorporate the right type of user research in an iterative design process. “And there is no substitute for being in the field with the customers, with the patients, with the clinicians, for that,” he said.

“We'll do our formative research to understand the user context and what their motivations and their goals are,” said Bucher, “and we will absolutely bring design thinking methodology into that. So not just usability testing once we have a prototype, but what we call desirability or usefulness testing, having somebody reflect to us, whether it's a product that they're interested in using, does this fit your lifestyle,” she concluded.

Niksch said that social determinants of health should be considered. “We've talked a lot about physical disability, physical limitations, but also I want to talk about assumptions that we make about language barriers, about ability to understand how data is presented on a digital health platform.” She also said that taking note that accessibility to connectivity is not equal everywhere is important.

Miller summed up by saying that part of the beauty of this iterative work is that it can identify unmet needs and new solutions can be found. “You might develop a whole new strategy or whole new device around solving a problem that you may never have thought of, and wouldn’t have done, if you had not had that kind of research available to you,” she concluded.

“Design & Product Development Principles for Connected Devices” and more sessions at MD&M BIOMEDigital are available to watch on demand through May 7. Registration assistance is available here.