Japan's Latest MHLW Guidelines Identifying Software as Medical Devices

The healthcare sector has witnessed a rapid surge in the development and market introduction of groundbreaking products and services, particularly those integrated with intelligent technologies. A notable instance of this trend is the growing integration of AI and machine learning technologies into medical devices aimed at diagnostics and testing. Concurrently, we're observing a proliferation of software applications and wearable gadgets designed to empower regular consumers in collecting health-related data and similar functions.

Naturally, the process of bringing innovative healthcare products and services to market must align with stringent healthcare regulations, including Japan's Pharmaceuticals and Medical Devices Act. The determination of whether a product falls under the category of a "medical device" has substantial ramifications for a company's regulatory approach, commercialization strategy, and the allocation of time and resources needed for a successful launch. Despite its importance, companies often grapple with discerning whether their innovative offerings will be classified as medical devices, largely due to the lack of clear precedents in this realm.

Addressing this challenge, the Ministry of Health, Labour and Welfare (MHLW) has issued the "Guidelines for determining whether a program is a medical device," aimed at assisting companies in evaluating whether the software will be categorized as a medical device. The Guidelines were most recently updated on March 31, 2023, to offer enhanced elucidation and direction (Yakuseiki-hatsu 0331-1, Yakuseikanma-hatsu 0331-4 on March 31, 2023).

The updated Guidelines furnish revised insights into the factors that bear weight in the assessment of whether software qualifies as a medical device. One key factor pertains to whether the software conducts disease diagnoses or indicates potential disease risks—an expanding sector of software products. Among these, certain products are devised for diagnosing or preventing diseases, whereas others offer insights into future disease risks to enhance overall health and well-being. Discerning the demarcation between software that does and does not fit the medical device category is a complex task. The revised Guidelines provide greater clarity regarding the considerations pivotal in making this determination, along with a visual aid for evaluating whether software "indicates a risk of disease" (referred to as "Disease Risk Indicating Software") and thereby qualifies as a medical device.

The stance of the MHLW is that any software delivering highly credible, "publicly known" medical or pharmaceutical information, or merely presenting excerpts from existing medical literature to users, does not meet the criteria for a medical device. This holds even if the software indicates potential disease presence or risk for the user. The Guidelines additionally elucidate that information is deemed "publicly known" when it holds scientific validation recognized within the spheres of medical, pharmaceutical, and nutritional science, and when the software's mere provision of this information doesn't automatically confer it with "publicly known" status.3 Software confined to delivering dependable, publicly known information to users avoids medical device classification because it refrains from independent medical analysis. In contrast, the Guidelines specify that products transcending this boundary—those incorporating novel algorithms or features—may indeed qualify as medical devices.

The Guidelines go on to outline that Disease Risk Indicating Software targeted at a "specific individual" might indeed be classified as a medical device. In simpler terms, if software specifies the disease risk for an individual, it might be deemed to perform disease diagnosis or prevention for that particular person. However, software that solely juxtaposes data supplied by an individual with statistical disease risk data from a particular population is not seen as a medical device since it doesn't focus on the individual's health.

Lastly, the Guidelines stipulate that software signaling a user's potential ongoing illness is likely to fall under the medical device ambit. This is particularly applicable when software deduces conceivable diseases the user might be grappling with, including their potential severity. Conversely, the Guidelines propose that software concentrating on healthy individuals and foretelling their future disease risks to "encourage health by adopting primary preventive measures" is improbable to be classified as a medical device. Nonetheless, there's a caveat—software concentrated solely on future disease risks could still be categorized as a medical device if it's intended for disease diagnosis, treatment, and prevention, or if such a purpose is explicitly indicated.

In conclusion, the 2023 updated Guidelines offer the most contemporary and pragmatic counsel on potentially classifying healthcare-related software as medical devices. This encompasses detailed insights into Disease Risk Indicating Software, grounded in the framework of Japanese legal precedents. Businesses embarking on developing and launching novel healthcare products and services should refer to these latest Guidelines to ascertain whether their offerings fall within the medical device scope in Japan.