Digital therapeutics from bench to bedside

The paper addresses how digital therapeutics (DTx)—a subcategory of digital medicine within digital health—can transition from research to routine clinical practice. It highlights ambiguity around DTx due to a lack of harmonized definitions, variable regulatory oversight, and uneven clinical evidence, distinguishing DTx from wellness apps by their evidence-based targeting of specific medical conditions (often chronic and neuropsychiatric) and from broader software as a medical device (SaMD) by their therapeutic intent. The objective is to map definitions, indications, commercialization, clinical trials, and regulations for DTx, identify challenges at the intersections of R&D, clinical validation, and regulation, and propose considerations for establishing DTx as standard medical practice.

The paper synthesizes prior definitions of DTx (e.g., early framing as evidence-based behavioral treatments delivered online; DTA’s definition emphasizing software-driven therapeutic interventions). It contrasts DTx with pharmacotherapy across development costs and timelines, manufacturing and distribution, trial pathways, side-effects, adherence, maintenance via updates, and post-market dynamics. It summarizes global definitional and regulatory heterogeneity, noting Korea’s government-level definition and most other jurisdictions treating DTx as SaMD without a distinct category. Prior commercialization has focused on conditions amenable to behavioral change (e.g., neuropsychiatric and chronic diseases), with expanding indications. The review also covers core technology elements (content such as CBT modules, games, VR; monitoring; communication), patient engagement and digital literacy requirements, and sociocultural factors potentially influencing efficacy and adoption.

The authors conducted a multi-pronged review (2010–2022): (1) Literature: Searched PubMed, IEEE, Google Scholar, ScienceDirect, and Web of Science using the string: (Digital Therapeutics OR DTx) AND (FDA OR SaMD OR MDR OR MDR AND (Digital Health OR Healthcare) AND (Smartphone OR Application). Non-journal articles were excluded; articles found via the string were reviewed, including those not in English. (2) Clinical trials: Identified via the same literature search, the Digital Therapeutics Alliance (DTA) product library, and ClinicalTrials.gov using the keyword “digital therapeutics.” Only trials with an NCT number and registered/presented between 2010 and 2022 were included. (3) Commercial products: Compiled from DTx-related literature, the FDA medical device database, and products with CE marks in Germany (DiGA) and Belgium, as well as listings on DTA, DiGA, and healthbelgium. (4) Regulations: Collected DTx-related regulatory and market documents from IMDRF member jurisdictions and other authorities using keywords “DTx” or “Digital therapeutics” OR “DTx” AND “regulation.” Documents that mentioned DTx explicitly in scope/keywords/main content were included; general SaMD-only documents were excluded. Data extraction focused on definitions, indications, trial characteristics, regulatory pathways, and commercialization status.

• Definitions and scope: No international consensus definition exists. Korea formally defines DTx as SaMD providing evidence-based therapeutic intervention for prevention/control/treatment. Most jurisdictions subsume DTx under SaMD without a dedicated category.
• Comparative characteristics: Relative to drugs, DTx have much lower development cost (<1% of pharmaceuticals) and shorter timelines (≈3–7 years vs ≈12 years), require no ongoing manufacturing, can be distributed directly via platforms, exhibit higher adherence (≈80% vs ≈50%), and are updatable post-market. Side effects are typically minor and device-related. Efficacy may depend on demographics, sociocultural and cognitive factors, and digital literacy. Barriers include engagement, prerequisites (device ownership, literacy), and data security.
• Clinical trials landscape: From screening 50,872 items, 78 DTx-related clinical trials were identified. Indications: psychiatric 31.1% (highest), poisoning 8.9%, cardiovascular 6.7%, with expansion to neurodegenerative (e.g., Parkinson’s), oncology, immunology, pain, autism spectrum disorders, schizophrenia, major depressive disorder, postoperative pain, lupus, and post-stroke dysarthria. Countries (registered): US 27 (60%), Switzerland 2 (4.4%), Finland 3 (6.1%), South Korea 2 (4.4%), Italy 4 (4.4%), Brazil 1 (2.2%), Malaysia 1 (2.2%), Singapore 1 (2.2%), Canada 1 (2.2%), Israel 1 (2.2%), Poland 1 (2.2%); 3 cases (6.7%) lacked ClinicalTrials.gov information. Sponsors: companies 24 (53.3%), pharmaceutical orgs 4 (8.9%), universities 7 (15.6%), hospitals 4 (8.9%), research institutes 3 (6.7%), medical consortiums 2 (4.4%), other 1 (2.9%). RCTs and parallel designs predominated; primary outcomes were mainly treatment, followed by research.
• Commercialization: Initial approvals focus on chronic and neuropsychiatric diseases (e.g., diabetes, COPD/asthma, addiction, insomnia, ADHD), with newer launches in IBS, migraine, orthopedic postop recovery, ear disorders, urinary incontinence, tinnitus, vaginismus. Many products cleared by FDA (510(k) and De Novo) and CE-marked; launches concentrated in the US and Europe (Germany, Belgium). Examples: reSET (CBT for substance use disorder), EndeavorRx (video game-based ADHD therapy), Propeller Health and ProAir Digihaler (inhaler-linked adherence platforms with reported reductions in inhaler use), Sleepio (web-based CBT-I). Sankey analyses indicate diversification of indications over time and company-led trials.
• Regulation: Within IMDRF jurisdictions, no harmonized DTx-specific framework; DTx generally regulated as SaMD under existing medical device laws. Korea issued multiple DTx-specific guidelines (definitions, submission requirements, and condition-specific evaluation designs for alcoholism, insomnia, nicotine use disorder, panic disorder). The US and EU pathways demonstrate that approvals are feasible via current device frameworks; temporary COVID-19 enforcement discretion facilitated psychiatric digital device access. Researchers call for clearer, harmonized standards, nomenclature (e.g., GMDN), and post-market requirements.

Findings underscore fragmentation in DTx definitions and regulatory treatment, contributing to uncertainty in development, evaluation, and adoption. The authors propose a practical definition—software that delivers evidence-based medical interventions for prevention, management, and treatment—and derive core requirements: ability to provide medical intervention, accessible software delivery, and evidence from systematic clinical trials. They advocate for international harmonization across terminology (potentially via IMDRF), performance standards (IEC/ISO), and condition-specific clinical protocols to streamline approvals and comparability. Given DTx dependence on patient engagement and digital literacy, demographic and sociocultural factors should be integrated into trial designs, usability, and regulatory evaluation, with equity-focused design to mitigate digital divides. While RCTs establish efficacy, real-world evidence (RWE) is vital to assess generalizability, long-term use, and safety, suggesting the need for stronger post-market surveillance and studies despite many DTx being moderate-risk. Regulatory education and guidance for nontraditional device manufacturers are also needed. Overall, aligning research, clinical validation, and regulation—while addressing engagement and access barriers—can accelerate safe and effective clinical integration of DTx.

The study maps the current DTx ecosystem—from definitions and characteristics to clinical trials, products, and regulation—and identifies key gaps hindering routine clinical use: lack of harmonized definitions/standards, variable and sometimes insufficient clinical evidence, and limited post-market data. It recommends international agreement on terminology and performance standards, standardized efficacy trial protocols where feasible, and incorporation of DTx terminology into global device nomenclatures. To ensure durable clinical adoption, the authors call for strengthened real-world evidence through coordinated efforts among researchers, manufacturers, and regulators, alongside technology and regulatory strategies to overcome engagement and access barriers. These steps can support DTx maturation from promising digital interventions to reliable components of mainstream medical practice.

• Data sources were restricted to publicly available information; in-development research, forthcoming trials, and draft regulations were not fully captured.
• Reporting bias toward successful implementations likely; failures and negative results were not systematically analyzed.
• Real-world effectiveness literature was largely excluded due to limited product-level access, constraining assessment of post-market performance.
• Reimbursement pathways were not addressed in depth; few mature models exist (e.g., DiGA in Germany, IAPT in the UK), and US legislation (Access to Prescription Digital Therapeutics Act) was not yet implemented, limiting comparative analysis.
• Few multinational and Asian clinical trials were identified, potentially limiting generalizability across regions and cultures.