China's digital therapeutics (DTx) business has vaulted into the limelight as a major new investment target of the healthcare sector, with a number of domestic companies joining the global Digital Therapeutics Alliance (DTA).

Digital therapeutics are changing the healthcare delivery system with evidence-based technologies that improve patient outcomes – and how companies position and successfully commercialise their products after regulatory approval has become one of the key issues to their development. This article summarises the major existing domestic regulatory policies and introduces digital therapeutics by analysing its definition, classification and registration approval procedure.


As defined on the DTA's official website, digital therapeutics is driven by high-quality software programs that provide patients with evidence-based therapeutic interventions to prevent, manage and treat diseases. Digital therapeutics can be used alone or in combination with medication, or in conjunction with other therapies to improve patient care and health status.

By function, digital therapeutics can be divided into three categories – prevention, management and treatment. Among domestic companies with the label "digital healthcare", respectively falling under the categories of prevention, management and treatment, 66Nao provides online electronic brain health cloud services; ClouDr provides monitoring and management services for diabetic patients; and the Recovery Plus app assists clinicians in cardiopulmonary rehabilitation training for patients.


At present, domestic regulatory policy for digital therapeutics is still in its infancy, and there are no clear regulatory rules. Considering that digital healthcare products are mostly in the form of standalone software, or software containing a medical device, in practice it is immediately necessary to confirm whether they are categorised as medical devices.

If such a product is defined as a medical device – based on intended use of the product combined with its processing objects, core functions and other factors – the more applicable concepts under the current guidelines are "mobile medical devices" and "artificial intelligence (AI) medical software".

Specifically, mobile medical devices refer to devices and/or software that use non-invasive mobile computing terminals to achieve one or more medical purposes. AI medical software refers to standalone software based on medical device data and AI technology to achieve its medical purposes, with core functions including medical device data processing, measurement, model calculation and analysis.

According to the Classification Catalogue for Medical Devices, classification of digital therapeutics products includes treatment planning software, image processing software, data processing software, decision support software, in vitro diagnosis software, and other products such as rehabilitation training software. The risk level can be comprehensively determined based on product description and intended use, and the regulatory provisions on class II/III medical devices apply. If such a product is not a medical device, the regulatory requirements are relatively simple. It is not required to comply with the relevant regulations on medical devices for R&D, testing, clinical trials, registration and production.


In the example of the Recovery Plus app (among the first batch of digital therapeutics products approved by China's authority of drug supervision), by assisting clinicians in guiding patients in cardiopulmonary rehabilitation training, it is categorised under above-mentioned rehabilitation training software, and is thus subject to regulatory provisions on class II medical devices.


For registration of class II/III medical devices, clinical evaluation is required in advance – unless the product is covered by the Catalogue of Medical Devices Exempted from Clinical Evaluation, which came into effect on 1 October 2021.

After a preliminary search, the rehabilitation training software to which the Recovery Plus app belongs was not included, which means this type of medical device should still be evaluated clinically. In addition, whether a product is required to go through clinical evaluation may be affected by its type of algorithm and application scenario, as well as the communication with competent authorities.


The registration process of digital therapeutics products is similar to that of ordinary medical devices; from sample testing, clinical evaluation, submission of application, technical review, administrative examination and approval to eventually obtaining a Registration Certificate for Medical Device.

In particular, digital therapeutics products should follow technical review guidelines for medical device software in the registration process. These mainly include: Guidelines for Technical Review of Medical Device Software Registration, the general guidelines for medical device software proposing requirements for software updates and versions; Guidelines for Technical Review of Mobile Medical Device Registration, focusing on risks arising from adoption of mobile computing technology in medical devices and control measures; Guidelines for Technical Review of Medical Device Cybersecurity Registration, setting out technical review requirements for medical device cybersecurity; Key Points for Reviewing Medical Device Software for In-depth Learning-Assisted Decision-making, applicable to registration application of medical device software for in-depth learning-assisted decision-making; and Guidelines for the Classification and Definition of Artificial Intelligence Medical Software Products, affirming that attribution to class II/III medical device regulation should be determined according to the maturity level of the algorithm in medical application, and whether it aids decision-making.

The following points should also be noted for digital therapeutics products:

In terms of research data, if the software uses mobile terminals, its cybersecurity capacity, display limitation, ambient light influence and battery capacity should be studied.

Digital therapeutics products may be used in combination with other medical devices, drugs and non-medical devices to achieve the same intended purpose, while research data proving the safety and effectiveness of such combined use should be provided.

As digital therapeutics products are evidence-based, non-clinical data such as evidence-based medical literature related to the product is also required to be submitted.

With big data and algorithm support, the digital therapeutics industry is moving full speed ahead. The authors believe regulators will soon make comprehensive improvements to their digital therapeutics oversight, and provide guidance to the flourishing business, in order to safeguard the lives and health of more patients.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.