South Korea is the only country to officially recognize digital therapeutics as a category. Nevertheless, the concept of software-based tools designed to manage or treat medical conditions has received enormous attention in the U.S. as proponents tout their potential to both widen access to care and improve patient outcomes. This optimism has driven significant financial investment in the sector, which ballooned to $3.4 billion in 2021.
However, recent setbacks and billions of dollars in bankruptcies in 2023 indicate that the field is stagnating. Though industry voices point to regulatory hurdles as the culprit, deeper structural challenges threaten the future of “software as a medical device” and other digital health technologies, as categorized by the Food and Drug Administration.
On the surface, the main impediment to digital therapeutics is the law. For an item or service to be covered by the Centers for Medicare and Medicaid Services, the Social Security Act requires it to have a benefit category. The act further mandates that items and services with a Medicare benefit category may only be covered if they are “reasonable and necessary for the diagnosis or treatment of illness or injury.” This criterion, though not explicitly defined, is considered to be more rigorous than the “safe and effective” standard that the FDA is required to use for a drug or device to be marketed.
As a novel innovation, digital therapeutics don’t neatly align with existing Medicare benefit categories. Some contend that they should be included in the durable medical equipment (DME) category and have criticized CMS for interpreting the definition of DME too narrowly. The DME argument was pursued by Pear Therapeutics, known for creating reSET — the first FDA-authorized digital therapeutic for substance use disorder. However, in June 2022, CMS denied Pear’s request, asserting that “digital therapies or computer software do not qualify as devices, equipment, or supplies” and thus do not fit within the DME benefit category. Less than a year after the CMS decision, Pear Therapeutics declared bankruptcy.
Creating a new benefit category for digital therapeutics requires legislation from Congress, such as the Access to Prescription Digital Therapeutics Act of 2023, which is currently under review. However, attempts to pass related legislation began as early as 2020 and have shown minimal progress. Few commercial payers cover digital therapeutics, highlighting deeper reservations about their utility.
But there are good reasons for both Congress and payers to be skeptical, at least at the moment. Concern regarding efficacy is one of the most important factors influencing coverage decisions. While proponents cite a wealth of peer-reviewed studies involving digital therapeutics, the field has received criticism for its low standard of evidence. One review of 14 meta-analyses found significant shortcomings across 145 trials, leading researchers to conclude there was a lack of convincing evidence to support mobile phone-based interventions for conditions like depression and nicotine use disorder. A 2023 review of 164 digital therapeutics studies identified from ClinicalTrials.gov identified low engagement as a chief barrier and found only a single study focused on basic-science understanding of mechanisms of action. No study was able to answer core questions such as the mechanism of action, optimization of the intervention, or dosing, which reflects the nascent state of the field.
Taken together, paying for today’s offerings could result in many prescriptions with limited patient adherence, raising costs for payers with minimal benefits for patients.
In the face of these concerns and no viable path to reimbursement, large players in the industry are experiencing setbacks. In September, Akili laid off 40% of its workforce, stating the company will focus on its direct-to-consumer business. In a further signal of the waning optimism of digital therapeutics, Biogen shuttered its 150-person digital health group, simultaneously ending its partnership with Apple.
The industry has two possible paths forward.
One approach involves conducting large, placebo-controlled randomized trials to clearly demonstrate evidence for digital therapeutics. If companies want their products to be covered under a pharmacy benefit model and paid for like traditional drugs, they should meet a comparable standard of evidence. Moving digital therapeutics away from medical devices and creating them with the rigorous evidence and regulatory pathways required of pharmaceuticals is a route that companies have not yet explored. This new generation of studies would include early phase investigation around preliminary efficacy, engagement, cultural appropriateness, and then later phase studies of more successful software versions that should be tested in rigorous digital placebo-controlled studies.
The risk in this approach is the cost of large, randomized trials and the possibility that digital therapeutics are not as differentiated as digital health companies want us to believe. Although taking this tack would require a lot of time, it will be expedited compared with biological drug discovery and testing.
Alternatively, another option is to reimagine the framework in which digital tools are incorporated into medical care.
Evidence shows that digital mental health interventions are more effective with human involvement, highlighting the importance of hybrid care models in which digital solutions are one part of a more comprehensive treatment plan. Evolving reimbursement frameworks, such as remote therapeutic monitoring, aim to pay for devices that are used to deliver internet-CBT and monitor patient response to treatment. Under this model, digital therapeutic companies would sell to providers, who would then submit for reimbursement by third-party payers, provided that certain monitoring requirements are met. Digital health companies would differentiate themselves based on features and consumer preferences.
This business model resembles a traditional software as a service approach more than the pharmaceutical model. While margins might be thinner, the sheer size of the potential customer base presents a promising financial opportunity. Given the importance of human support, coaching models like Digital Navigators will serve as a foundational platform from which more tailored support offerings can evolve. Challenges related to workforce availability for these new roles will likely spur the expansion of novel service providers such as peer-support specialists. Future research will discern whether clinical improvements are driven by human support, software, or both, with the likely result that there is a valuable synergy worthy of investment.
Although the term “digital therapeutics” suggests clinical utility, both the implied benefits and the name itself remain ill-defined and largely unrecognized globally. The recent high-profile setbacks and failures in the digital therapeutic space may reflect an overall market correction. The device pathway, with its lower bar for regulatory approval, may be too low a bar for the high expectations set by both the public and investors for this new field. Investing in new business models that demand an approach as rigorous as traditional pharmaceutical development will help develop the next generation of tools that may deliver on the promise. In the short term, investing in business models that use existing, and often simple tools-enhanced by human support-will likely deliver value by achieving small but meaningful outcomes at scale.
G. Luke Hartstein is a fourth-year psychiatry resident at Beth Israel Deaconess Medical Center. John Torous is an assistant professor of psychiatry at Harvard Medical School and staff psychiatrist at Beth Israel Deaconess Medical Center. He is the scientific advisor for Precision Mental Wellness, a company he holds equity in.