FDA and CluePoints Extend Existing Collaboration to Include AI/ML for Quality Assessment

FDA and CluePoints have announced a new 3-year extension to their long-standing collaboration. The duo will be working to further enhance CluePoints’ SMART software.¹

Under the original cooperative research and development agreement (CRADA) between the FDA and CluePoints from 2016,² new software was developed and CluePoints’ existing software was enhanced to produce a ranked list of anomalous sites to help prioritize site inspection(s) for FDA inspectors.

Due to the large number of product submissions, the FDA can only inspect a limited number of clinical trial sites. The agency relies on site inspections and whistleblowers to detect fraud and other forms of misconduct. The determination of which sites end up being inspected can involve the judgement of inspectors, suggestions by reviewers, and CDER’s risk-based site selection tool. Since the original CRADA, CluePoints and the FDA have worked together on prioritizing these inspections.

According to a press release, under the newly proposed CRADA, the FDA and CluePoints will focus on two primary objectives:

• The first objective is to develop and enhance CluePoints’ SMART software to address a broader range of regulatory issues and concerns. The CRADA will leverage date/time data which takes on greater significance with decentralized clinical trials and the increasing use of electronic clinical outcome assessment and electronic patient reported outcomes technologies. Further developments will be added to improve the detection of duplicate patients. The proposed CRADA will explore how artificial intelligence and machine learning algorithms can further support anomaly detection. Finally, further research will be carried out to explore the moderators of treatment effect and develop a software solution that can be deployed within the FDA environment.
• The second objective is to improve and enhance how SMART software and the CluePoints monitoring platform may be adapted to better support FDA processes related to anomaly detection, review and follow-up, as well as site selection for FDA inspections.

“FDA and its stakeholders have a vested interest in ensuring the integrity of clinical trial data and the safety of participants while clinical research is being conducted,” the press release reads. “Clinical research misconduct, such as fabrication or omission of data in reporting study results, puts all individuals in that trial at risk of harm. Fraud and other forms of misconduct impairs FDA’s goal to safeguard and promote public health by jeopardizing the accuracy of data provided to the agency.”

References

1. FDA and CluePoints Sign New 3 Year Cooperative Research and Development Agreement to Assess Data Quality using Statistical Modelling and Machine Learning. News release. June 4, 2024. Accessed June 4, 2024. https://www.prnewswire.com/news-releases/fda-and-cluepoints-sign-new-3-year-cooperative-research-and-development-agreement-to-assess-data-quality-using-statistical-modelling-and-machine-learning-302164842.html

2. FDA Signs Agreement With CluePoints To Explore A Data-Driven Approach To Quality Oversight In Clinical Trials. News release. October 12, 2016. Accessed June 4, 2024. https://cluepoints.com/fda/