Interventional trial registrations doubled between 2004 and 2025, reaching more than 10,000 annually, and the average number of eligibility criteria per protocol rose more than 60 percent over the two decades prior. That arithmetic lands on EDC operations as a compounding problem: more studies, more data sources, more sites, and more manual touchpoints at every layer. The question MedNet Solutions raises in a recent analysis of its cubeCDMS platform is whether AI’s primary value in EDC is speed, or something more useful: fewer tasks entering the queue in the first place.
The distinction matters operationally. Automating existing workflows still assumes those workflows exist. CRScube’s approach targets the front end of the problem. Its AI-led study setup reads a CRF specification and configures the EDC automatically, including trial-specific pages and edit checks, with human review and approval before anything is finalized. Configuration that previously took days collapses to minutes. Across a portfolio, that multiplier is where the capacity gain accumulates. The EHR-to-EDC functionality follows the same logic: instead of relying on API connections and site IT cooperation that can delay a site’s activation by weeks, the tool works directly inside the data-capture workflow. Site staff direct the system to relevant fields in an electronic health record; AI populates the corresponding CRF entries for review. FDA guidance on the use of EHR data in clinical investigations does not require EHR systems themselves to comply with 21 CFR Part 11, which means integration-free approaches can fit within the existing regulatory framework without requiring sponsors to negotiate technical access with every new site.
Two features address avoidable downstream work. An interactive edit check surfaces potential data issues before a CRF page is saved, stopping a query workflow before it starts. A WHODrug search assist suggests standardized medication terms at the point of entry rather than flagging non-standard entries after the fact. One prevented query is trivial; thousands prevented across a large site network removes an entire reconciliation layer that would otherwise require monitor time, data manager review, and site follow-up.
The native-integration argument runs underneath all of this. Adding AI through standalone tools or external vendors trades one kind of complexity for another, and operations teams that already manage multiple systems feel that cost acutely. Embedding these capabilities directly in cubeCDMS sidesteps that problem. The metric to watch as sponsors evaluate this approach is not build time in isolation, but whether query volume per site actually drops as study complexity increases, because that ratio is what determines whether the scaling equation has really changed.
Source link: https://www.mednetsolutions.com/blog/ai-in-edc-scaling-without-exploding/
Moe Alsumidaie, MBA, MSF, is founder and Chief Editor of Vanguard Publications, which publishes Clinical Trial Vanguard, Pharma Vanguard and BullScope, and Head of Research at CliniBiz. He has two decades in clinical trial operations and data science, with earlier roles at Genentech, Abbott Vascular and Stanford University Medical Center, and is a guest lecturer in clinical trial sciences at Rutgers University.
