As technology advances and the need for precision care increases, the link between medical software and AI has opened up new opportunities for innovation in healthcare. In one of our previous assignments, with a company specialized in the development of products that present medical information based on tissue images through AI and ML (machine learning), we had the opportunity to support a project where a medical device was about to enter the global market. The company faced some challenges, not least because they wanted to secure an approval from the FDA (Food and Drug Administration) for the US market.
Challenges of lack of information & limited knowledge
The technical file, which contains detailed technical information about the medical device, needed a major update from the In Vitro Diagnostic Medical Devices Directive (IVDD) to the In Vitro Diagnostic Medical Devices Regulation (IVDR) with increased requirements for detailed information on areas such as intended use, product classification, risk analysis, software-related documentation, AI, cybersecurity and clinical studies.
The company also faced the challenge of limited experience with documentation specific to software development and IVDR. In addition, products with AI and machine learning posed another challenge in itself due to the fact that only a few similar medical devices had been reviewed for approval by the FDA (these are called Predicate Devices). Since the field was relatively new and few products had been approved for the market, the process at the FDA becomes even more demanding.
Analyses and strategies delivered the desired results
To address these challenges, we made a fresh start with a thorough review of existing documentation. To facilitate the communication with the FDA, a clear description for the implementation of the FDA submission was created, where we managed to identify a suitable Predicate Device. A comprehensive GAP analysis was carried out to update the technical file in accordance with the IVDR. Extra focus was put on meeting the requirements especially in the area of cybersecurity from both the Notified Body (for the European market) and the FDA.
An overall strategy, based on GAP analysis, was included in the final delivery and greatly facilitated the company, which needed some further guidance to meet the regulatory requirements in order to reach the market with its product. Thanks to the implementation of development processes (in accordance with current guidelines and standards for AI and machine learning), a good basis for product development was ensured. In addition, the company was able to submit a pre-submission to the FDA to get formal input for the registration of the product on the US market. The technical documentation was updated to meet the requirements for global product registrations.
Summing up
We can see that the regulatory requirements in medical software challenge manufacturers to raise the bar for compliance, not least to ensure both cyber and patient safety. This requires not only diligence at every step of the process, but also a proactive approach to adapt to changes.
Is your company facing similar challenges? Contact us to learn more about our work in MedTech and IVD, where the focus is on tomorrow's cutting-edge products, their benefits and patient safety.
