Prescription Digital Therapeutics: Unmet Needs
Healthcare providers comment on what clinicians would like to see, in terms of data and education, that would help justify the prescribing of digital therapeutics for behavior-driven conditions.
Megan Coder, PharmD, MBA: Dr Moukaddam, regarding the provider side, what is your experience, and what you would want to see in an experience with patients? Given that we are talking about products that are used in the context of care with the clinician, I’d be interested to find out what you would like to have as a part of that engagement with your patients and this product?
Nidal Moukaddam, MD, PhD: I would want something that provides me with objective data that are scalable. I would like to be able to try it out and use it to know that it’s beneficial and the scalability comes in when I say as a specialist, “This is great, but I can give it to everybody who’s a PCP [primary care physician] and everybody who is doing this on a larger scale.” That is important. The main issues in mental health, insomnia, and addiction are that what we’re going for is not measured very well. Psychiatry still follows a categorical approach. If I say you’re depressed, it might mean you look sad, but technically what it means is that you meet the criteria for a certain number of symptoms in a duration.
Digital therapeutics do not give you that. Digital therapeutics give you a wealth of information about how much you move, sleep duration, and sleep quality. Instead, digital therapeutics give information on how many times you’ve turned your phone on or off. This is the kind of thing I can use to track your activity and your sleep cycle. What we do in our research is we use privacy preserving methods. Even so, I don’t know who you talked to, but I can still draw a social map. Therefore, I can see how this helps predict the relapse when you start calling random people and I notice this is not your usual pattern.
The field of digital therapeutics should give me something objective that would be useful. To achieve that, I’m going to stick to the example of apps and wearables because this is what most people are used to. Let’s assume your phone, just like your wristband that measures your activity. These devices are all things that are very intimately paired with your lifestyle. This device can give you 2 inlets. It can ask you how you’re feeling, which is called ecological momentary assessment. Additionally, it can ask you if you cough in your sleep, what you’ve eaten, if you’ve had a problem physically or a symptom. Or it doesn’t have to ask you anything, and it just detects your patterns. That pattern detection is a prolific and powerful tool, which is what we need as clinicians. We need that analysis to bridge the gap of what we think of as symptoms. Am I making sense?
Megan Coder, PharmD, MBA: Yes. Going back to some of these questions in general, what is a digital therapeutic, and how does it relate to the broader ecosystem? My background is in pharmacy, so sometimes I’ll hear patients ask, “Is it something that will just track my medication adherence?” The answer is the software in a digital therapeutic does deliver the medical intervention, but it could be paired with a wearable or a fitness tracker or a monitoring or diagnostic component. What I love about the industry is that digital therapeutics are delivering an active intervention to a patient, yet they’re working in conjunction with remote patient monitoring and telehealth and all other products to enable holistic care. That’s what you’re getting at—this notion of holistic care and something that provides actual data to decide on, so you don’t feel as if you’re working in a vacuum.
Nidal Moukaddam, MD, PhD: That is called adjusted time intervention. It takes a typical AI [artificial intelligence] algorithm 10 to 14 days to learn your patterns. Notably, the people who do this are so smart, and I can’t follow 10% of what they’re saying, but they give me wonderful algorithms. We can track patterns, and the idea of adjusted time intervention is to take this pattern and say, “Today you’re not doing what you usually do. Today you did not do your usual 2 miles of walking.” From there, you can link it to whatever condition you think this patient has. Are you sick? Are you depressed? Are you not feeling well? Is your asthma acting up? You have the detection component.
The diagnostic component obviously is a lot trickier, and you must have a lot of regulatory oversight for that because we don’t want to jump from detection to diagnosis, which is a big area. First, we have to get those algorithms and validate them. For mental health and addiction specifically, the field is going to have to see a much larger switch from categorical diagnosis. Specifically, do you have this dimensional diagnosis? The dimensional would consist of rating your depression only, or specifically rating your appetite. We’re not talking about the symptom cluster but rather 1 symptom.
Megan Coder, PharmD, MBA: Dr Luo, what type of education would be needed to really enable you to have confidence to use these types of tools?
John Luo, MD: Papers published and FDA clearance or approval. When I was a junior faculty member and the PalmPilot came out, 1 of the ways I increased adoption—and these were not digital therapeutics or anything close to that—is I would teach a course. In 1999, 100 psychiatrists would vie for a seat to learn how to use their PalmPilot, just to use it in providing basic access to information.
Finding a way to reach psychiatrists and help them be comfortable prescribing it or using it or showing it to their patients reminds me that like electronic health records, much of medical education has been outsourced to these third-party vendors, which create videos to help the patient with lumbar punctures for example.
You certainly can have the same vendor teaching digital therapeutics. You probably don’t want it on YouTube, but those are the ways providers will become more comfortable. Patients are way ahead of us because they’re eager. They really want support, especially because access to mental health providers is somewhat limited.
