Research coordinators spend 12 to 15 hours per week per study copying data that already exists in a hospital EHR into a separate EDC system. That number, cited by Mednet, points to something the EDC market’s projected growth from $3.2 billion to $7.1 billion by 2030 does not resolve on its own: a bigger market does not fix a workflow where more than half of trial data gets entered twice.
The delay cost compounds that burden. Manual EHR-to-EDC transcription holds data in limbo for 14 to 30 days before sponsors and CROs see it, which means query management and oversight decisions run on information that is already weeks old. Mednet’s answer, built through its partnership with CRScube, skips the usual fix of adding another integration vendor. Instead, the AI intake feature sits inside cubeCDMS itself. A coordinator opens the patient’s EHR as usual, and the AI assistant reads what is on screen, matches it against the relevant eCRF fields, and copies the data across. No separate deployment, no trial-specific mapping project, no third-party middleware to maintain.
The design choice matters more than it might appear. Sites increasingly factor technology friction into which trials they agree to run, and sponsors that reduce administrative load gain a real recruiting edge. A 2023 meta-analysis in PubMed put the pooled error rate for manual medical record abstraction at 6.57 percent, a figure that compounds across every data point a coordinator transcribes by hand. Eliminating the transcription step attacks error at the source rather than catching it through downstream queries. That shifts the data management team’s time from correcting routine entry mistakes to reviewing data that actually needs a human judgment call.
The FDA has not issued a standalone guidance for AI-driven EDC capture, so regulatory characterization of this kind of screen-based intake remains part of the broader conversation around AI in clinical investigations. Sponsors evaluating cubeCDMS will need to satisfy themselves on source data verification requirements specific to their protocols. The practical test will be whether sites running studies on the platform see the transcription hours actually drop, and whether cleaner first-pass data translates to fewer query cycles before database lock.
Source link: https://www.mednetsolutions.com/blog/ai-emr-intake-breaking-down-data-silos-in-clinical-research/
Moe Alsumidaie is Chief Editor of The Clinical Trial Vanguard. Moe holds decades of experience in the clinical trials industry. Moe also serves as Head of Research at CliniBiz and Chief Data Scientist at Annex Clinical Corporation.
