Digital Therapeutics in Viet Nam: What They Are, What Evidence Should Support Them, and How They May Fit the Current Regulatory Landscape
POLICY & REGULATORY INSIGHTS
Digital Therapeutics in Viet Nam: What They Are, What Evidence Should Support Them, and How They May Fit the Current Regulatory Landscape
Last reviewed: 24 September 2026
A growing number of healthcare products are described as “digital therapeutics.” The term can sound intuitive. If an application helps someone manage diabetes, improve sleep, reduce anxiety or follow a rehabilitation program, it may seem reasonable to call it a digital therapy.
But the distinction is more important than it first appears. A wellness application, medication reminder, patient-education platform, remote-monitoring system and digital therapeutic can all support health, yet they do not necessarily perform the same function, require the same evidence or create the same level of clinical and regulatory risk.
For healthcare organizations in Viet Nam, this is becoming a practical question rather than simply a matter of terminology. As software takes a larger role in disease management and treatment, hospitals, clinics, technology companies and investors will increasingly need to understand what the software actually does, what therapeutic claim is being made, what evidence supports that claim, what regulatory framework may apply and whether the intervention can be integrated safely into clinical care.
What digital therapeutics actually are
ISO/TR 11147:2023 describes digital therapeutics, commonly abbreviated as DTx, as a category of health software that generates and delivers medical interventions based on clinical evidence, has demonstrable positive therapeutic impacts on patient health and produces real-world outcomes. It also describes the relationship between digital therapeutics and other digital-health concepts, including medical devices, Software as a Medical Device and other digital health technologies.[1]
The Digital Therapeutics Alliance similarly defines DTx as health software intended to treat or alleviate a disease, disorder, condition or injury by generating and delivering a medical intervention with a demonstrable positive therapeutic impact. Its principles also emphasize safety, efficacy, quality, patient centricity, privacy, security, appropriate regulatory review and ongoing clinical impact.[2]
The important concept is the therapeutic intervention. Software is not a digital therapeutic simply because patients use it, because it contains medical information or because it improves access to healthcare. The software itself needs to play an active therapeutic role.
That role can take different forms. A digital therapeutic might deliver structured cognitive behavioral therapy, guide disease-specific behavioral treatment, provide software-directed rehabilitation, use biofeedback, personalize a therapeutic program or support another intervention through clinically defined treatment logic. It may be used independently or alongside medication, a medical device or clinician-delivered treatment. It also does not need to use artificial intelligence. A rules-based software program can fall within the digital-therapeutics concept when the software itself generates and delivers the therapeutic intervention and the therapeutic claim is appropriately supported.
A digital therapeutic is not simply another health app
Many digital products sit close to digital therapeutics without actually being digital therapeutics. A medication-reminder application may improve adherence but does not necessarily deliver treatment itself. A remote patient monitoring platform may collect blood pressure or glucose measurements without generating a therapeutic intervention. A telemedicine platform connects a patient with a healthcare professional, but treatment is generally delivered by the professional rather than by the platform. Diagnostic software may identify disease or calculate risk without treating it, while a wellness application may support exercise, nutrition, relaxation or sleep without making a therapeutic claim related to a particular disease or condition.
The same product can, however, contain several functions. A diabetes platform might combine education, remote monitoring, medication reminders and a software-delivered therapeutic intervention within one ecosystem. Evidence and regulatory questions may therefore need to be considered function by function rather than assuming that the entire platform has a single identity.
This is one reason terminology should follow intended purpose and actual function, not marketing language. Calling an application a “digital therapeutic” does not make it one. Conversely, avoiding the term does not necessarily remove medical-device or healthcare obligations if the software is actually being used to deliver a medical intervention.
Evidence is not an optional feature of digital therapeutics
Evidence is one of the characteristics that distinguish a credible digital therapeutic from a conventional health or wellness application. If software claims to treat or alleviate a disease, the evidence supporting that claim should be proportionate to the clinical significance and potential risk of the intervention.
A useful framework comes from the International Medical Device Regulators Forum guidance on clinical evaluation of Software as a Medical Device. It identifies three related elements: valid clinical association, whether there is a valid relationship between the software output and the targeted clinical condition; analytical validation, whether the software correctly processes its inputs to generate accurate, reliable and precise outputs; and clinical validation, whether use of those outputs achieves the intended purpose in the target population and clinical context.[3]
Applied to a digital therapeutic, the questions become practical. Is there a scientifically credible basis for the therapeutic mechanism? Does the software deliver or generate the intended intervention consistently and correctly? Most importantly, when patients actually use it, does it produce the clinical benefit being claimed?
A technically excellent application can still fail the final question. Good user-interface design, reliable coding and accurate algorithms are important, but none of them independently demonstrates therapeutic effectiveness.
Evidence should begin with the intended use
Before asking whether a digital therapeutic has “good evidence,” a healthcare organization needs to know exactly what the product claims to do. The intended population matters. Evidence in adults cannot automatically be transferred to adolescents, while evidence in people with uncomplicated hypertension may not establish effectiveness in patients with multiple chronic conditions. Evidence for supporting disease self-management is also not necessarily evidence for replacing clinician-delivered treatment.
The intended outcome matters as well. Application engagement, completion of educational modules or number of logins may demonstrate product use, but they are not automatically therapeutic outcomes. A product claiming to improve diabetes control should be evaluated against clinically meaningful outcomes relevant to diabetes. A product claiming to treat insomnia needs outcomes that appropriately reflect sleep and patient benefit. A rehabilitation intervention should demonstrate meaningful improvement in the function it claims to restore or improve.
The product version matters too. Digital products can change much faster than traditional therapeutic interventions. If the treatment logic, algorithm, therapeutic content, personalization model or decision rules change substantially, evidence generated using an earlier version may no longer fully support the product currently being offered.
Clinical effectiveness requires more than a before-and-after improvement
Patients may improve after using an application for many reasons, including natural variation in disease, medication changes, increased clinical attention, regression to the mean or behavioral changes unrelated to the software. A before-and-after improvement alone does not establish that the digital intervention caused the benefit.
For technologies intended to treat specific conditions, the NICE Evidence Standards Framework recommends one or more high-quality interventional studies supporting the claimed benefit and showing relevant clinical or patient outcomes. Randomized controlled trials are preferred where appropriate, although high-quality comparative real-world designs may also be acceptable in suitable circumstances.[4]
A randomized controlled trial can provide particularly strong evidence of causality and will often be appropriate for a digital therapeutic, but not every product or every stage of evaluation requires exactly the same study design. Pragmatic trials, comparative observational studies and real-world studies can answer additional questions. What matters is whether the design is capable of credibly testing the claim being made.
Digital interventions also create methodological challenges that traditional drug trials do not always face. Participants often know which intervention they are receiving, engagement can decline over time, software can change during development, and treatment effects may depend heavily on how frequently or correctly patients use the product. These realities do not justify lowering the evidence standard. They need to be anticipated in study design and reported transparently.
Evidence for the category is not evidence for an individual product
There is an expanding research base for digital therapeutics, but DTx should not be treated as if it were one therapy. A 2024 analysis identified 280 eligible digital-therapeutics trials registered on ClinicalTrials.gov. Among completed trials, 67.5 percent used randomized controlled designs. Therapeutic areas and study methods varied considerably, however, and the authors highlighted continuing inconsistency in clinical-trial design and outcome selection.[5]
A 2026 npj Digital Medicine study further illustrates why product-level review matters. Researchers identified 45 digital therapeutics with United States Food and Drug Administration authorization decisions between 2019 and 2022 and located 436 studies associated with them. The evidence base varied substantially. A pivotal randomized controlled trial was identified for 46.7 percent of the products, no studies were identified for 11.1 percent, and 53.3 percent had no publicly available study results before their initial authorization.[6]
These findings do not mean that digital therapeutics generally lack evidence. They show why neither the DTx label nor regulatory status should substitute for examining the evidence supporting the particular product, indication and version being considered.
There are also encouraging findings in specific clinical areas. A 2025 systematic review and meta-analysis involving 19 studies and 3,264 participants with diabetes found reductions in glycated hemoglobin and several other metabolic measures associated with DTx interventions, with particularly favorable findings for glycated hemoglobin among participants with type 2 diabetes.[7] A separate meta-analysis of 15 randomized controlled trials involving 3,789 adults with hypertension found modest improvements in blood pressure and several related risk factors, while also calling for larger trials evaluating the same products.[8] These findings are useful signals for those clinical areas, but they should not be generalized to every digital therapeutic or patient population.
Safety includes usability, engagement and the possibility of harm
A digital therapeutic cannot be evaluated only by asking whether the software crashes. Poor usability can become a clinical risk if patients misunderstand instructions, select the wrong option or stop using an intervention before receiving sufficient therapeutic exposure. Harm can also arise from inappropriate recommendations, delayed escalation to professional care, excessive reassurance, poorly managed deterioration or adverse psychological effects.
Engagement therefore needs to be interpreted carefully. High engagement may support treatment delivery, but engagement itself is not the therapeutic objective. A healthcare organization should understand how much exposure appears necessary to achieve benefit, how many participants completed the intervention in clinical studies, why patients discontinued it and whether effectiveness was maintained outside highly supported research environments.
Human factors are particularly important when users include older adults, people with disabilities, patients with limited health literacy or people using the technology while experiencing illness or psychological distress. Accessibility and usability should therefore be considered part of therapeutic quality and safety, not simply interface design.
Evidence generation should continue after launch
Software does not stop evolving after launch, so evidence generation should not stop either. Real-world evidence can help determine whether benefits observed in controlled trials persist when the product reaches more diverse populations with fewer research supports. It can also reveal differences in adherence, adverse events, healthcare utilization, access and the effects of subsequent software updates.
The 2026 npj Digital Medicine study illustrates the importance of looking at the entire evidence lifecycle. Some authorized digital therapeutics had substantial evidence spanning development, pivotal trials, real-world use and subsequent product modifications, while others had much more limited evidence.[6]
For healthcare organizations, version control should therefore be part of evidence governance. If version 3.2 of an intervention was evaluated in a clinical trial, a substantial change to the therapeutic algorithm in version 4.0 should not automatically be assumed to carry exactly the same evidence base. The organization should understand what changed and whether the change could alter safety, effectiveness or the applicability of earlier evidence.
Evidence generated overseas may not fully transfer to Viet Nam
An intervention may be clinically effective in another country and still require careful evaluation before implementation in Viet Nam. Language is only the most obvious issue. Digital therapeutics can depend on behavioral assumptions, dietary patterns, health literacy, medication availability, clinical pathways, socioeconomic circumstances, family involvement and interaction with healthcare professionals. Translating an English interface into Vietnamese does not by itself validate the therapeutic intervention for Vietnamese patients.
The healthcare context also matters. A product studied in a system where patients have frequent access to specialist teams and integrated electronic health records may operate differently where care pathways, workforce models or integration capabilities differ. Smartphone access may be widespread, but digital literacy, connectivity, device quality and willingness to share sensitive information can vary substantially among populations.
For an imported digital therapeutic, the question should not be whether every international trial must be repeated in Viet Nam. A more useful question is whether the available evidence can reasonably be transferred to the proposed Vietnamese population and care setting, and whether targeted local validation or a structured implementation study is needed to address important remaining uncertainty.
Viet Nam does not currently have a standalone digital therapeutics regulatory category
As of this review, the principal Vietnamese healthcare, medical-device, artificial intelligence and data regulations examined for this article do not establish a standalone statutory definition or dedicated authorization pathway specifically named “digital therapeutics.”
That does not mean digital therapeutics sit outside regulation. It means the applicable pathway needs to be determined from the product's intended purpose, therapeutic claims, functions, risk, architecture, data processing and the way it is used in healthcare.
This distinction is important for companies entering Viet Nam. Regulatory authorization, certification or conformity status obtained for a digital therapeutic in the United States, European Union, Republic of Korea, Australia or another jurisdiction can provide useful supporting information, but it does not automatically determine the product's Vietnamese regulatory status.
Medical-device regulation will often be part of the analysis
Viet Nam's current medical-device framework is principally based on Decree No. 98/2021/NĐ-CP, as amended and consolidated in Ministry of Health Consolidated Document No. 08/VBHN-BYT dated 6 March 2026.[9] The statutory definition expressly includes software where the specified criteria are met, including software intended for purposes such as diagnosis, prevention, monitoring, treatment or alleviation of disease or injury.
Because a genuine digital therapeutic uses software to generate and deliver a medical intervention, the medical-device framework will often be an important part of the regulatory analysis. It should not, however, be converted into a blanket statement that every product marketed internationally as a digital therapeutic automatically follows exactly the same Vietnamese medical-device pathway.
There is an important software-specific nuance in the current framework. Article 2 expressly includes software within the definition of a medical device when the applicable criteria are satisfied. At the same time, Article 3(8) provides that requirements concerning classification, issuance of circulation numbers and declaration of eligibility for trading do not apply to “phần mềm (software) sử dụng cho thiết bị y tế”, which can be rendered in English as software used for medical devices.[9]
The precise application of these provisions to a particular standalone therapeutic software product should therefore not be inferred from the DTx label alone. The product's intended purpose, architecture, functionality and relationship with any other medical device need to be examined on their own facts. Healthcare organizations and technology companies should avoid both extremes: assuming that all therapeutic software is exempt from medical-device requirements, or assuming that every digital therapeutic follows exactly the same pathway as a conventional physical medical device.
Where classification requirements apply, Viet Nam uses a risk-based medical-device framework. Circular No. 24/2026/TT-BYT, effective from 1 July 2026, regulates risk determination and management measures for products and goods that are medical devices and amends provisions of Circular No. 05/2022/TT-BYT. The consolidated rules are reflected in Consolidated Document No. 12/2026/VBHN-TT-BYT dated 10 July 2026.[10]
Artificial intelligence is a separate question
Digital therapeutics do not need artificial intelligence to function. Many evidence-based therapeutic interventions use deterministic treatment logic, structured behavioral protocols or other software functions without artificial intelligence.
If artificial intelligence is incorporated, however, another regulatory layer may become relevant. Viet Nam's Law on Artificial Intelligence No. 134/2025/QH15 has been effective since 1 March 2026, while Decree No. 142/2026/NĐ-CP has been effective since 1 May 2026.[11,12] Under the current framework, the provider is responsible for classifying an artificial intelligence system before it is put into use. Systems classified as medium or high risk require a classification dossier and notification in accordance with the applicable requirements.[11,12]
A digital therapeutic should not be assumed to be a high-risk artificial intelligence system simply because it is used in healthcare. Under Decision No. 33/2026/QĐ-TTg, effective from 15 August 2026, the healthcare section of the current high-risk list contains two categories focused on artificial intelligence supporting surgery and robotic surgery. Other healthcare artificial intelligence systems still need to be assessed under the broader risk-classification framework according to their actual characteristics and use.[13]
Medical-device regulation and artificial intelligence regulation address different questions. Depending on the product, both may be relevant. Where artificial intelligence is embedded within therapeutic software, organizations should therefore consider the regulatory requirements applicable to the product as well as those applicable to the artificial intelligence system itself.
Digital treatment does not remove professional responsibility
A digital therapeutic can be used directly by a patient, recommended by a clinician or incorporated into a wider clinical pathway. These models are not legally identical.
Where healthcare professionals use a digital therapeutic as part of medical examination or treatment, responsibilities established under Viet Nam's Law on Medical Examination and Treatment remain relevant. The presence of software does not, by itself, shift responsibilities that the law places on practitioners or healthcare facilities.[14]
The distinction from telemedicine is also important. A patient can use a digital therapeutic at home without participating in a remote consultation. Conversely, a clinician may review information produced by the product during remote medical examination or treatment, in which case the separate Vietnamese requirements governing remote healthcare may also become relevant.
Healthcare organizations should therefore design the clinical pathway around the actual model of care. They need to understand when the software operates independently, when a healthcare professional reviews progress, what findings require escalation, who is expected to respond and what patients should do if their condition worsens.
Personal and health data should be considered from the beginning
Digital therapeutics can generate particularly rich longitudinal datasets. Depending on the intervention, these may include symptoms, diagnoses, medication information, behavioral patterns, emotional states, sleep, physical activity, treatment responses and interactions with therapeutic content.
Viet Nam's Law on Personal Data Protection No. 91/2025/QH15 has been effective since 1 January 2026. Article 26 requires the data subject's consent during the collection and processing of personal data relating to health information, except in the circumstances specified in Article 19(1). It also requires organizations and individuals developing health applications to comply fully with applicable personal-data protection requirements.[15] Decree No. 356/2025/NĐ-CP provides further implementation requirements.[16]
The Law on Data No. 60/2024/QH15 and Decree No. 102/2025/NĐ-CP on management of health data, both effective from 1 July 2025, add another important layer to the governance of health information.[17,18]
A digital-therapeutics company should therefore be able to explain much more than whether it has a privacy policy. Healthcare organizations need to understand which data are collected, which information is actually necessary for the therapeutic purpose, the respective roles of the parties, which subcontractors can access the data, where they are stored, how long they are retained, whether they may be reused for analytics, product development or artificial intelligence training, and how patients can exercise their applicable rights.
Foreign cloud infrastructure deserves particular attention. Article 17 of Decree No. 356/2025/NĐ-CP treats several situations as cross-border personal-data transfers, including transferring personal data collected or stored in Viet Nam to servers overseas, storing or processing such data through cloud services of a foreign provider, and transferring data collected in Viet Nam to an overseas platform for further processing. Applicable impact-assessment requirements and statutory exemptions therefore need to be determined from the actual data architecture rather than simply from the vendor's headquarters or brand.[16]
Cybersecurity is part of therapeutic safety
When software itself delivers treatment, a cybersecurity incident can become more than a confidentiality problem. A compromised account could expose sensitive information, a corrupted treatment rule could result in an inappropriate intervention, an unavailable service could interrupt treatment, and an unauthorized software modification could change what patients receive.
Viet Nam's Law on Cybersecurity No. 116/2025/QH15 has been effective since 1 July 2026. The implementation framework was further developed in August 2026 through Decree No. 331/2026/NĐ-CP on cybersecurity protection for information systems and Decree No. 333/2026/NĐ-CP detailing provisions and measures for implementation of the Law on Cybersecurity. Decree No. 330/2026/NĐ-CP establishes the administrative-sanctions framework for cybersecurity and personal-data protection. These three decrees took effect on 19 August 2026.[19–22]
For healthcare organizations, cybersecurity review should therefore be connected to clinical risk. Authentication, authorization, encryption, vulnerability management, software updates, application programming interfaces, audit trails, incident response, backup, recovery and vendor access all matter, but their importance increases when system failure can interrupt or alter a therapeutic intervention.
Clinical records and interoperability still matter
A digital therapeutic may operate on a patient's phone, but clinically important information should not disappear into an isolated vendor ecosystem. If clinicians rely on information from the product when making treatment decisions, the organization should determine what needs to become part of the clinical record.
This does not mean every screen tap, behavioral interaction or raw application event belongs in the electronic medical record. It means that clinically meaningful information, such as treatment progress, relevant alerts, significant outcomes or information used to support a clinical decision, should be identifiable and documented in a way that supports continuity of care.
Circular No. 13/2025/TT-BYT has governed electronic medical-record implementation since 21 July 2025 and establishes implementation requirements and timelines for healthcare facilities within its scope.[23]
Interoperability also matters before scale. A pilot involving a few dozen patients can tolerate manual exports and separate dashboards. A program involving hundreds or thousands of patients cannot rely indefinitely on clinicians switching between disconnected systems. Healthcare organizations should therefore consider patient identification, structured-data exchange, medical-record integration, access to treatment summaries and data portability before entering a long-term contract.
Reimbursement should not be assumed
Regulatory status, evidence of clinical effectiveness and reimbursement are three different questions. A product may have strong evidence and an appropriate regulatory position without having a defined mechanism for payment through Viet Nam's health-insurance system.
As of this review, the national health-insurance framework examined for this article does not establish a dedicated reimbursement pathway specifically for products described as digital therapeutics. Existing provisions concerning remote medical examination and treatment or payment for medicines used in remote care should not automatically be interpreted as reimbursement mechanisms for therapeutic software. The current consolidated Law on Health Insurance was issued as Consolidated Document No. 40/VBHN-VPQH on 16 March 2026.[24]
Healthcare organizations and technology companies should therefore evaluate financing separately from the clinical and regulatory assessment. Depending on the product and implementation model, early adoption may involve healthcare-organization procurement, research or pilot arrangements, patient-funded services or other financing approaches permitted under applicable rules and contracting arrangements. Foreign reimbursement or regulatory recognition should not be assumed to create health-insurance coverage in Viet Nam.
Procurement should evaluate the therapy, not just the application
A digital therapeutic should not be evaluated in the same way as ordinary enterprise software. Procurement needs to consider the clinical indication, intended population, therapeutic mechanism, precise product version, evidence supporting the claimed outcomes, safety profile, Vietnamese regulatory position, data architecture, cybersecurity controls and integration requirements.
Evaluation should also be multidisciplinary. Clinical teams need to understand the therapeutic value and safety implications, quality and patient-safety teams need to understand how the intervention fits into care processes, information-technology and data-governance teams need to assess infrastructure and data risks, and procurement needs to ensure that contractual obligations reflect the clinical importance of the product.
The vendor should be able to explain what happens when the software changes. Who determines whether an update is clinically significant? Is additional validation performed? Can the healthcare organization continue using the previous version while an important update is assessed? Can treatment logic change automatically? How are patients and clinicians informed?
Commercial contracts should address these issues alongside technical support, uptime, data access and portability, security incidents, subcontractors, service termination and business continuity. Once software becomes part of treatment, continuity of service is no longer simply an information-technology issue.
Implementation should start with a clinical problem, not with a technology
For Viet Nam, the most useful starting point is unlikely to be a hospital announcing that it wants “a digital therapeutics program.” A better starting point is a defined clinical problem.
A healthcare organization might identify a population in which existing care leaves an important gap, such as patients who need structured therapeutic support between visits, people who struggle to sustain a treatment-related behavior or individuals whose access to an established intervention is limited. The organization can then ask whether a software-delivered therapeutic intervention is an appropriate way to address that gap.
The product should be evaluated against the current care pathway rather than against having no digital product at all. If a simpler intervention already delivers good outcomes at lower cost and complexity, a digital therapeutic should be able to demonstrate what additional clinical or operational value it creates.
Initial implementation should usually be controlled in scope. A pilot can assess patient selection, onboarding, adherence, safety events, clinician workload, technical problems, escalation pathways, clinical outcomes and equity before wider deployment. Success criteria should be defined before the pilot begins rather than declaring success simply because patients downloaded the application or completed a specified number of sessions.
What a credible digital therapeutic proposition should demonstrate
A credible digital therapeutic should tell a coherent story from therapeutic claim to patient outcome. The disease or condition should be clearly defined, and the software should have an identifiable therapeutic mechanism. The intervention delivered by the current product version should correspond sufficiently to the intervention that was clinically evaluated, and the evidence should match the intended population and claim.
Safety, usability, accessibility, privacy and cybersecurity should have been addressed systematically. The Vietnamese regulatory position should be understood rather than inferred from foreign approval. If healthcare professionals are involved, their roles and escalation responsibilities should be clear. There should also be a practical plan for data governance, clinical integration, product updates and continued evaluation after implementation.
This standard is deliberately higher than asking whether an application looks professional or contains scientifically accurate information. Digital therapeutics ask patients and healthcare professionals to rely on software as part of treatment. That creates a correspondingly higher obligation to demonstrate that the software is suitable for that role.
Where digital therapeutics may fit in Viet Nam
Digital therapeutics could become an important part of Digital Medicine in Viet Nam, particularly where healthcare needs extend beyond episodic visits and evidence-based interventions can be delivered repeatedly outside conventional healthcare settings. Chronic disease management, behavioral health and rehabilitation are among the areas worth exploring, but adoption should follow evidence and clinical need rather than technology trends.
Software can extend therapeutic support into patients' everyday lives and potentially reach people whom conventional models do not serve well. The same scalability, however, means that a poorly validated intervention can also be distributed widely and quickly. Viet Nam therefore does not need to adopt digital therapeutics simply because the category is growing internationally. What matters is developing the capability to distinguish credible therapeutic software from ordinary digital health products, evaluate the evidence rigorously, understand the applicable regulatory framework and test implementation in real Vietnamese healthcare settings.
The central question is not simply “Can treatment be delivered through software?” In some circumstances, we already know that it can. The more useful question is “For this particular patient population, does this particular software deliver a clinically meaningful therapeutic benefit, with sufficient evidence, safety, governance and regulatory clarity to justify its use in care?”
That is the standard that can help digital therapeutics become meaningful Digital Medicine rather than simply another category of health technology.
References and further reading
International Organization for Standardization. ISO/TR 11147:2023 Health informatics — Personalized digital health — Digital therapeutics health software systems. 2023.
Digital Therapeutics Alliance. What Is a DTx? Understanding Digital Therapeutics. Accessed September 2026.
International Medical Device Regulators Forum. Software as a Medical Device (SaMD): Clinical Evaluation. IMDRF/SaMD WG/N41FINAL:2017.
National Institute for Health and Care Excellence. Evidence Standards Framework for Digital Health Technologies. Published 2018, updated August 2022.
Yao H, Liao Z, Zhang X, et al. A comprehensive survey of the clinical trial landscape on digital therapeutics. Heliyon. 2024;10(16):e36115.
Amyx M, Alebouyeh F, Phi NTT, Ravaud P, Tran VT, et al. Mapping the evidence supporting FDA-authorized digital therapeutics, before and after authorization. npj Digital Medicine. 2026;9:618.
Li M, Chen Y, Chen X, Yao H, You L. Impact of Digital Therapeutics for the Management of Adult Patients With Diabetes: Systematic Review and Meta-Analysis of Randomized Controlled Trials. Journal of Medical Internet Research. 2025;27:e70428.
Liu L, Guo J, Jiao X, You L. The effect of digital therapeutics intervention on improving hypertension management in adults: a meta-analysis of randomized controlled trial. Hypertension Research. 2025;48:456–469.
Government of Viet Nam. Decree No. 98/2021/NĐ-CP on medical-device management, as amended. See Ministry of Health Consolidated Document No. 08/VBHN-BYT dated 6 March 2026.
Ministry of Health of Viet Nam. Circular No. 24/2026/TT-BYT on determination of risk levels and management measures for products and goods that are medical devices and amendments to Circular No. 05/2022/TT-BYT. Effective 1 July 2026. See also Consolidated Document No. 12/2026/VBHN-TT-BYT dated 10 July 2026.
National Assembly of Viet Nam. Law on Artificial Intelligence No. 134/2025/QH15. Effective 1 March 2026.
Government of Viet Nam. Decree No. 142/2026/NĐ-CP detailing provisions and measures for implementation of the Law on Artificial Intelligence. Effective 1 May 2026.
Prime Minister of Viet Nam. Decision No. 33/2026/QĐ-TTg promulgating the List of High-Risk Artificial Intelligence Systems. Effective 15 August 2026.
National Assembly of Viet Nam. Law on Medical Examination and Treatment No. 15/2023/QH15. Effective 1 January 2024.
National Assembly of Viet Nam. Law on Personal Data Protection No. 91/2025/QH15. Effective 1 January 2026.
Government of Viet Nam. Decree No. 356/2025/NĐ-CP detailing provisions and measures for implementation of the Law on Personal Data Protection. Effective 1 January 2026.
National Assembly of Viet Nam. Law on Data No. 60/2024/QH15. Effective 1 July 2025.
Government of Viet Nam. Decree No. 102/2025/NĐ-CP on Management of Health Data. Effective 1 July 2025.
National Assembly of Viet Nam. Law on Cybersecurity No. 116/2025/QH15. Effective 1 July 2026.
Government of Viet Nam. Decree No. 331/2026/NĐ-CP on cybersecurity protection for information systems. Effective 19 August 2026.
Government of Viet Nam. Decree No. 333/2026/NĐ-CP detailing provisions and measures for implementation of the Law on Cybersecurity. Effective 19 August 2026.
Government of Viet Nam. Decree No. 330/2026/NĐ-CP on administrative sanctions in cybersecurity and personal-data protection. Effective 19 August 2026.
Ministry of Health of Viet Nam. Circular No. 13/2025/TT-BYT providing guidance on implementation of electronic medical records. Effective 21 July 2025.
National Assembly of Viet Nam. Law on Health Insurance No. 25/2008/QH12, as amended. See Consolidated Document No. 40/VBHN-VPQH dated 16 March 2026.
Ministry of Health of Viet Nam. Circular No. 37/2024/TT-BYT, particularly Article 14 concerning health-insurance payment for medicines used in remote medical examination and treatment, remote support and remote consultation. Effective 1 January 2025.
Regulatory and educational note
This article provides a general overview of digital therapeutics in Viet Nam as of 24 September 2026. The term “digital therapeutics” is internationally used to describe health software that generates and delivers evidence-based therapeutic interventions, but the principal Vietnamese legal instruments reviewed for this article do not currently establish a standalone regulatory category or dedicated market-authorization pathway specifically named Digital Therapeutics.
The Vietnamese regulatory status of a particular product therefore needs to be assessed from its intended purpose, therapeutic claims, functions, risk, architecture and actual use. Software may meet the Vietnamese definition of a medical device when the statutory criteria are satisfied, while the current medical-device framework also contains specific provisions concerning software used for medical devices. Product-specific assessment is therefore important, and foreign regulatory status should not automatically be treated as determining the Vietnamese pathway.
Where artificial intelligence is incorporated into a digital therapeutic, the artificial intelligence framework should be assessed separately. A healthcare use case does not, by itself, mean that an artificial intelligence system falls within Viet Nam's current high-risk list.
Similarly, evidence that a digital therapeutic is clinically effective does not, by itself, establish regulatory authorization, health-insurance reimbursement or suitability for implementation in a particular healthcare organization. These are related but separate questions.
This article is for educational purposes and does not constitute legal, regulatory or clinical advice.