Financial interests and evidence in public comments on the FDA framework for modifications to artificial intelligence/machine learning-based medical devices
Smith JA., Abhari R., Hussain Z., Heneghan C., Collins G., Carr A.
The US Food and Drug Administration (FDA) has proposed a regulatory framework for modifications to artificial intelligence/machine learning-based software as a medical device (SaMD). We analysed the 125 non-duplicate and non-spam public comments on the proposed framework and found that industry and other parties with a potential financial interest contributed a minimum of 63% of comments and that the absence of a conflict of interest could be confirmed in only 8%. Despite the FDA's stated commitment to using scientific evidence in the rule-making process, the vast majority of submitted comments (86%) did not cite any scientific literature, just 4% cited a systematic review or meta-analysis, and no comments indicated whether a systematic process was used to identify relevant literature. We recommend that the FDA requires disclosure of conflicts of interest in comments in the future. As the framework continues to be modified, engagement from non-conflicted parties and better use of scientific evidence may help to ensure the development of a regulatory system that best promotes patient interests.