Dr. Lyle Berkowitz on What Health Tech Can Learn from Game Developers

Human-centered design’s important role in health’s future

Dr. Lyle Berkowitz is not your everyday doctor.

While no stranger to the challenges routinely faced by physicians in their everyday schedule, as a practicing MD, FACP, FHIMSS, he’s also taken an active role in embracing the possibilities of technological innovation from outside healthcare as CEO of virtual care platform KeyCare, and acts as a regular media contributor. Dr. Berkowitz recently participated in the R/GA Live event “An Outsider View Into Healthcare,” where he spoke enthusiastically about the possibilities opened up by tech innovations making their way into health. 

“We don’t have a shortage of physicians – we have a shortage of using them efficiently,” he said, pointing to the way ChatGPT can be used for writing prior-authorization letters to free up their time for working with patients as one example. “We have to use technology to make sure everyone is working at the height of their license.”

Dr. Berkowitz also stressed that outside innovators seeking to enter the healthcare field need to have a better understanding of the complexities of the system, and that tech innovation in health needs to be grounded in human-centered design. 

We caught up with Dr. Berkowitz to gain further insight into these topics, what it would look like for doctors to truly work “at the height of their license,” and more. 

This interview has been edited for clarity and length.

FutureVision: In the recent R/GA Live event, you said that health practitioners could benefit from better understanding human-centered design. Could you elaborate on that point? How have you seen these benefits demonstrated, and where are these elements currently missing where they could improve experiences and/or outcomes for patients? 

Dr. Lyle Berkowitz: Well to clarify, I’d say that teams developing new products or services would benefit from understanding human-centered design. One of the traditional problems in health tech has been that a tech/product person asks a doctor or nurse what product they want and the provider explains how they do it on paper, and then the product person tries to replicate that – which is not a good way to design a product. 

Using human-centered design, they would observe the provider at work, ask them to explain what they are doing and what problem they want to solve, etc. to get a better sense of the types of solutions they can create, as opposed to trying to computerize a paper-based process. In other words, designers need to ask the right questions, do deep observation, and share empathy with providers to come up with better ideas. The analogy is with gaming: programmers didn’t simply replicate a board game – they created amazing interactive 3D worlds. How can we do that in healthcare?

Similarly, could you elaborate on how some outside parties interested in the space have made missteps because of a lack of understanding around healthcare? What would you suggest should be prerequisites for entering health spheres in terms of subject matter knowledge? Where would you suggest tech or design innovators start when looking to enter the space with a limited knowledge of healthcare? 

One of the biggest mistakes made is not understanding how incentives work in healthcare. Outside healthcare, you have a two-party system: a buyer and a seller, usually.  In healthcare, you have three distinct parties: payor, provider and patient – each with their own costs and incentives. Furthermore, we don’t have one consistent system – for example, we have both fee-for-service and value-based care models. 

These three questions always need to be aligned: Who uses the product? Who benefits from the use of the product, and how (through saved time, or better care, for example)? And who pays for the product? So a product that improves patient outcomes and decreases costs to the payor, but costs the physician extra time – and you want to charge the provider – is not well aligned.  

Physician comp is really a leadership choice, but it can feel complex – due to cultural, financial and legal issues to navigate. There are consultants that focus on this now. And as you would expect, it hurts physician adoption if you ask them to support something that can negatively impact their revenues.

A product that improves patient outcomes and decreases costs to the payor, but costs the physician extra time – and you want to charge the provider – is not well aligned.

One of the most visible proliferations in health tech has been in consumer wearable devices tracking a variety of biometric info. Do you see any potential cause for concern in over-testing as an unintended result?

The major concern is that it will cause more false positives – which then need to be further evaluated and waste time.

What would it look like for practitioners to be truly working at the height of their license without being encumbered by burdens of the current healthcare system? What benefits would this translate to for patients and providers? 

The providers would be head of a large tech-enabled team, so that they could oversee or manage a large patient panel size, but only need to actually see a small percent of them to help with initial diagnosis and treatment plan changes. All preventive care and chronic monitoring would be done by the team – with escalations as needed. In that situation, the provider would be paid on panel size and quality, not on how many patients they see and RVUs [Relative Value Units in medical billing] they bill. 

What do you think are the most promising uses of automation going forward, both for practitioners and for patients?

If I was starting something, I’d look at provider (doctor/nurse) pain points that take up time and have little to no revenue, and try to automate as much as possible. 

Two obvious areas are “admina-strivia” items… such as filling out paperwork, as well as  answering routine patient questions via the inbox. We are already seeing how AI like ChatGPT can start helping with those things.