Medical Device Safety Throughout the Product Lifecycle: Regulatory Frameworks, Digital Transformation, and the Role of Artificial Intelligence in Devices and Monitoring.

Abstract

Medical devices are a mainstay in modern healthcare, supporting the diagnosis, monitoring, treatment, and management of acute and chronic conditions. Rapid technological innovation, increasing software dependence, growing connectivity, and the emergence of artificial intelligence (AI)-enabled systems have expanded the traditional boundaries of device safety. This narrative review examines medical device safety across the product lifecycle, from early design and preclinical development through clinical investigation, regulatory review, postmarket surveillance (PMS), software change management, cybersecurity, and AI-enabled device oversight. It examines recent regulatory approaches and implications of modern technological advances. A structured literature search was conducted using PubMed and ScienceDirect for relevant publications from 2016 to 2026, as well as an additional literature search focused on clinical research topics for the date range of 2013 to 2026. Findings were supplemented by regulatory guidance documents, industry white papers, and key sources from the FDA, EMA, and Cochrane Library. Authors screened, reviewed, and summarized the publications collectively to use as a basis for the discussion. The review highlights that although drug and device safety share common principles, including adverse event (AE) monitoring, serious AE reporting, benefit-risk assessment, and signal management, device safety requires a broader socio-technical framework. Device-related harm may arise from product malfunction, device deficiency, use error, operator dependence, procedural factors, environmental conditions, maintenance practices, or workflow failures. Therefore, robust safety oversight requires integration of risk management, biocompatibility, usability engineering, quality systems, clinical investigation standards, complaint handling, and PMS. The findings also emphasize progressive transition from retrospective and passive safety reporting toward more proactive, real-time, and lifecycle-based evidence generation. Real-world data, active surveillance, decentralized trial models, digital health technologies, and AI-assisted signal detection are reshaping how safety risks are identified and managed. At the same time, software updates, cybersecurity vulnerabilities, and AI/ML-enabled device behavior introduce new risks that may not be fully captured during premarket evaluation. These developments require continuous monitoring, disciplined change control, human factors assessment, and stronger international regulatory alignment. Overall, medical device safety should be approached as a dynamic lifecycle discipline rather than a static premarket requirement. Future safety frameworks will depend on harmonized global reporting systems, stronger postmarket evidence generation, integration of cybersecurity into benefit-risk assessment, adoption of Good Machine Learning Practice, and continued attention to human factors in increasingly complex clinical environments.