1. Scope
- Intended Use and Exclusions: This guidance applies to home-use software medical devices (SaMD) intended to detect signs of disease using data collected from general-purpose digital devices and encourage users to seek medical consultation. Products intended for physician diagnosis, physician-directed disease management, symptom relief, or those requiring dedicated medical devices are outside the scope of this guidance.
- Intended Users: Disease detection SaMD is intended for healthy individuals or those with undiagnosed conditions seeking guidance on whether to seek medical attention, as well as diagnosed patients monitoring symptoms independently of physician instructions.
2. Product Description and Key Considerations
- Clinical Intention: It is critical to define the product as a home-use tool for detecting signs of disease rather than providing a diagnosis. Applicants should explain its intended clinical role, expected benefits, and its relationship to existing diagnostic or therapeutic approaches, healthcare pathways, and also the risks associated with false-positive and false-negative results.
- Design Concept: The design concept should translate clinical intent into technical specifications that can be readily understood by non-expert users.
- Software typically runs on general-purpose devices or cloud platforms, requiring a clear definition of its regulatory scope relative to non-medical functions.
- Key requirements include defining input data, target signs, and analytical principles, such as specific algorithms or AI models.
- Since users lack medical expertise, output results should be clear and non-misleading.
- Risk management should involve safeguards like pre-use activation to restrict access to appropriate populations and functions that assess the measurement environment.
- The design should provide concrete usage instructions, such as required posture or physical condition, to ensure reliable performance in a home setting.
- Predicate Device: Applicants should provide a thorough comparison between the predicate device(s) and the device under review, clearly explaining their similarities and differences.
3. Evaluation Dossier
Based on 2. Product Description and Key Considerations, the efficacy and safety of the device should be fully evaluated. The essential requirements are:
- Clinical performance
- Basic performance
- Software development lifecycle
- Cybersecurity
- Usability engineering
Regarding clinical performance, in addition to demonstrating clinical effectiveness, the risk of misinterpretation (false-positive/false-negative) should be evaluated and appropriate risk mitigation measures should be implemented. In addition, because there are no non-clinical models that can accurately replicate home-use environments, clinical trials involving human subjects are generally necessary.
4. Points to Consider for Study Design in Clinical Evaluation
- Test Samples and Subjects: Clinical evaluation for home-use disease detection software should ensure equivalence between test samples and product under review. If the general-purpose digital device, software functions, or software version used in testing differs from that of the product under review, these differences should be clearly identified, and the applicability of the test results to the product under review should be justified. Similarly, any differences between study populations and intended populations in clinical practice should be assessed, and the validity of study populations should be explained.
- Ground Truth Labels: “Ground Truth Labels” should be established based on events that have a clinically evident association with signs of the target disease (e.g., atrial fibrillation (AF) identified from ECG, or arrhythmias identified from heart rate data).
- Evaluation Items: Sensitivity and specificity are generally recommended evaluation metrics. If positive predictive value (PPV) is used, its dependence on disease prevalence should be justified, and false-negative performance should also be evaluated.
Please note that additional evaluation may be required depending on the specific technology used in the product. If there are any questions regarding the product, such as its clinical intention or clinical study protocol design, consultation with PMDA is recommended.
Understanding a new market often requires both local knowledge and regional experience. With our Japan local offices and professional team in Osaka & Yokohama and 26+ years of experience across Asia Pacific, Qualtech supports medical device companies in navigating market regulatory requirements and evaluating growth opportunities. Contact us to discuss your plans for the Japan region.