AI and Robust Scientific Foundations to Optimize Clinical Trial Operations

Ahead of DPHARM 2026, we speak to Marty Magazzolo, Chief Commercial Officer from CluePoints about how AI-led capabilities are helping optimize clinical trials – and the new products they’ll be demonstrating at this year’s conference.

Optimizing clinical operations and AI readiness are key themes at this year’s DPHARM conference. What is CluePoints doing in this area?

At CluePoints we are focused on turning artificial intelligence into human intelligence to transform a key pressure point in clinical operations – data management. Clinical data management has traditionally been fragmented by design, as well as separate systems, teams and workflows. That is why problems are identified late, reviewed multiple times, and ultimately acted upon too slowly.

We are setting a simpler standard. Only meaningful data issues should be reviewed, and they should be reviewed once by the right person, using the right data, at the right time. It is about shifting from human in the loop to human in the lead. The science does the searching, the expert does the deciding.

Clinical trials have never been short of data. What they have been short of is the confidence to act on it early enough to drive the right outcome. By detecting risks earlier and ensuring data quality we can run better, more efficient trials.

Can you give us specific examples of how technology is being used to create future-ready clinical trials?

Three of the most resource-intensive workflows in clinical development are query management, medical and safety review, and medical coding. These are the workflows where fragmented tools, broad manual checks and disconnected review cycles most often delay decisions and put data integrity at risk.

Earlier this month we officially launched three new clinical data review capabilities – Intelligent Query Detection (IQD), Medical & Safety Review (MSR) and Intelligent Medical Coding (IMC). All three were co-innovated in close collaboration with a top 10 pharmaceutical company and are now live in ongoing studies, including mega-studies.

We shaped, tested and validated these capabilities in production alongside data management, medical and coding teams, against real portfolio pressure. The teams contributed directly to the design of IQD’s user-defined scenarios and to MSR’s change-based review model – the two capabilities we expect to have the most immediate effect on how sponsors organize clinical data review.

IQD is designed to replace broad, manual listing review with targeted, evidence-backed query generation. Can you tell us more about that?

IQD puts hyper-efficient execution into the hands of the study team. Through user-defined scenarios, a data manager can ask any operational or clinical question. There is no code, no change request, no waiting on a vendor release cycle. The system applies that logic continuously and only returns records where the answer suggests something needs attention.

This is the point at which a data manager’s own study knowledge becomes a repeatable, automated check and detection logic becomes a study team’s capability. By raising high-precision queries earlier, IQD results in reduced site burden, faster, higher-quality responses and an accelerated path from Last Patient Last Visit to database lock.

Can you tell us mote about MSR’s change-based review model?

In many organizations, medical and safety review relies on spreadsheets, static listings and manual reconciliation leading to rework and handoffs. MSR transforms this process into a continuous, patient-centric decision workflow where reviewers are guided through live study updates proactively and move from a signal to a contextualized query inside the same validated environment.

Instead of burdensome, error-prone mapping, MSR unifies data from disparate sources. Instead of a stack of listings, it presents a consolidated clinical picture of a patient. Instead of relying on retrospective reconstruction, review activity is captured continuously, building audit-ready evidence as the work happens. MSR is system-agnostic and available for both ongoing and new studies, with deployment measured in days rather than months.

Clinical trials are becoming increasingly complex. How does a tool like IMC help to overcome that?

As clinical programs grow in volume and complexity, traditional medical coding workflows reach practical limits. There is a 70–80% accuracy ceiling and 20–30% of terms require manual review, rework, and coordination overhead.

IMC is an agentic application rather than another coding interface. It applies advanced deep learning (DL) to generate MedDRA and WHODrug coding suggestions, delivering up to 99% accuracy for MedDRA and reducing manual coding effort by approximately 50%. IMC learns from an organization’s own validated historical decisions, so institutional coding knowledge is retained across studies, teams and dictionary up-versions rather than resetting each time.

How can people find out more about your new clinical data review capabilities?

We will be hosting on-site demonstrations of IQD, MSR and IMC at DPHARM® 2026. Come see us at Booth #121 for further information.

How do these launches build on CluePoints’ work in risk-based quality management (RBQM)?

We built our reputation on applying peer-reviewed science to clinical oversight, an approach now enacted in ICH E6(R3) and deployed by CluePoints across thousands of studies globally. Risk-based monitoring has become the industry standard and IQD, MSR and IMC are built on that same rigorous and adaptable scientific foundation.

How can RBQM itself help optimize clinical operations?

Earlier this year we worked alongside the Tufts Center for the Study of Drug Development (CSDD) to publish peer-reviewed evidence quantifying the financial value of RBQM. The study addresses a long-standing gap in the industry, modelling financial value across clinical development phases.

RBQM-enabled trials were associated with 8%-19% reductions in clinical phase durations. At the clinical trial level, estimated financial returns ranged from $3.2M in phase 1 to $18.9M in phase 3. At the development-program level, Expected Net Present Value (eNPV) gains ranged from $3.8M in phase 1 to $13.8M in phase 3. The largest contributor to financial value was time savings. These results really reinforce the strategic value of proactive, data-driven oversight beyond monitoring-cost reductions alone.