Digital therapeutics (DTx) in primary care
The debate over DTx product definitions has been ongoing since their arrival on the market less than a decade ago. In brief, DTx products use algorithms instead of molecules to deliver clinically-validated therapeutic effects. Robust evidence and regulation are the main ingredients. What makes this conversation even more interesting is that many roads can lead to the same ambiguous destination. Bluestar by Welldoc was the first mobile prescription therapy via the 510(k) pathway for diabetes management for adults. Pear Therapeutics introduced reSET for substance use disorder via a newly created De Novo pathway after requesting the FDA to regulate them. Nightware, developed to treat nightmares, made its debut via Breakthrough Device designation, a program that can help companies get through either the 510(k) or the De Novo pathway quicker or with more guidance from the FDA (h/t Brian Dolan for helping me understand this nuance). Generally, the timeline for the 510(k) can be a couple of months while the De Novo pathway can take up to one and a half years (Figure 6).
In a recent editorial, Chief Clinical Innovation Officer at CVS Caremark calls for more consensus and less obscurity in identifying true DTx products and more consistent regulatory oversight. Organizations like Digital Therapeutic Alliance founded in 2017 are rising to the challenge to help all stakeholders navigate the DTx ecosystem. The 2021 IQVIA report further differentiates DTx and digital care (DC) products based on breadth of clinical indication (page 14). The following algorithm (somewhat) demystifies the identity crisis for digital interventions.
So what do the DTx products actually do? The 2021 Deloitte report has a nice illustration of use cases across prevention, management, and treatment spectra of DTx.
Segmenting the DTx market can be even less clear and full of outliers that represent the growing pains of providing medical care in novel ways. I found Blue Matter’s framework most useful in differentiating prescription digital therapeutics (PTD) from non-prescription digital therapeutics (NDT).
The 2021 IQVIA report estimates approximately 50 DTx products in use with nearly double that amount in early stages of development. My favorite visual from the report combines DTx product indications, delivery platforms, and market entry pathways into one graphic.

For those who think less in graphics and more in timelines, you might like the analysis from Blue Matter that provides a bird’s eye view of past and future DTx products.

The most clinically relevant illustration is below and depicts the strength and outcomes of scientific evidence that undergirds the DTx movement. The purple box in the top right corner mirrors the number and variety of conditions an average fee-for-service primary care physician sees in a single day. Is this a threat or an opportunity for primary care?

My take is that DTx products are net positive with a few caveats. DTx products offer increased armamentum of treatment options for our patients and enables us to have the kind of “therapeutic reach” that our current health systems do not allow for. Being able to hyper-personalize therapy and adjust algorithms in real time is the biggest promise for a new way of taking care of patients. Having access to clinical insights, not just data, on a population level will make scaling ourselves immensely easier, especially in times of worsening primary care physician shortages. Knowing real-time adherence to treatment plans and interval outcomes redefines the traditional “wait and watch” approach in primary care. For this to work, DTx data has to be integrated into clinical workflow within the electronic medical record (EMR). Poorly understood chronic disease processes are driven by behavioral variables and require high-touch care. DTx products can really be a game changer and make asynchronous primary care more desirable for all stakeholders.
The “do no harm” principle guides our careers in medicine. Physicians who go into primary care specialize in people, not organ systems. For us, medicine is about relationships, and relationships are about trust. Patients trust us to care for them in their most vulnerable moments. It is clear why DTx products need to be vetted for us to trust them. I am most cautious about two things: safeguarding health data and sustainability. Data is the new currency in health care and DTx products hold the potential to change the power imbalance of health data ownership. Secondly, chronic illnesses often require chronic care and chronic care requires long-term commitment. How can we preserve the fidelity of patient-doctor relationship in the emerging triad of patient-DTx-doctor? DTx products are not immune to short horizon investments of venture capital and private equity worlds. If a DTx product has excellent outcomes and high net promoter scores, but no longer generates the same margin, will it survive in the market? Dealing with the aftermath will be punted to primary care. There is not yet a mechanism for “DTx Lipitor” to become “atorvastatin.” Doing due diligence on sustainability cannot be ignored.
When I prescribe medications, especially for mental health disorders, I think about their mechanisms of action. For example, fluoxetine (Prozac) is a selective serotonin reuptake inhibitor that changes serotonin levels in the brain to treat depression. Similarly, I imagine thinking about “algorithms of action” of DTx products to select the right combination of psychosocial interventions, such as cognitive behavioral therapy (CBT), peer counseling, or biofeedback.
My choice of medication also depends on what the patient in front of me can actually afford, and not just what’s recommended by the latest specialty society guidelines. If DTx products are provided as a medical benefit, their price would reflect the cost of care in the traditional settings (approximately $200 per month). Multiply by 10 if it’s a pharmacy benefit and divide by 10 if sold direct-to-consumer. Current price of the ADHD therapy video game EndeavorRx is $99 for a three-month prescription and Somryst sleep therapy cost $899 for the same duration in the not so distant past. Digital formularies have been emerging since 2019, but none offer PDT. Notable examples of such formularies include Express Scripts and CVS Caremark who curate a vetted list of “digital solutions” for payers. Interestingly, some health plans explicitly state that they are not paying for PDTs due to “insufficient evidence in the published peer-reviewed literature to support their effectiveness” with annotations for each reviewed PDT (Aetna and Premera Blue Cross). The 2021 HealthXL - Exits & Outcomes report highlights four categories of PDT companies (example): Rx-only (Akili), Rx-only but not because of DTx component (Welldoc), non-Rx but use Rx distribution channel (SilverCloud Health), non-Rx with plans to become Rx-only (Kaia Health).

Much remains to be decided about DTx products, but it is certain their traction and success in the marketplace won’t be possible without support and commitment from primary care physicians. The most important practical question that needs to be answered next is “How do we evaluate and select the right DTx product for the right patient?” Perhaps it will be a mix of platforms like Digital.Health and Xealth.
For a deeper dive on the DTx pipeline, check out the Databases section of our XPC HealthTech Index.
For what it's worth, I think that starting with a use case is far better than a definition (ive been a member of DTA, participated in AMCPs forums, etc etc) and I am convinced that what we have here is a positioning issue not a definitional one. If we have clarity around setting / site of care, stage of disease (1st line, second line intervention etc) it will help resolve a lot of the confusion. As part of this, saying something like "you can try X instead of Y" makes a clear case for subsitition. Many DTx products can be tried prior to starting a pharmacological approach for example.