Novartis Restructures Clinical Ops to Scale Trial Innovation Beyond Pilots

Novartis has overhauled how it approaches clinical trial innovation — moving from fragmented, trial-by-trial pilots to a structured operating model designed to scale proven solutions across its entire portfolio. Scott Askin, Portfolio Innovation Director at Novartis Global Clinical Operations, presented the model at the Clinical Trial Innovation Summit 2025 in Basel, offering a detailed look at how the team operates 18 months in.

The Problem: Fragmented Innovation That Didn’t Scale

Before the restructure, Novartis ran innovation the way most large pharma companies do. Teams would identify a new technology, find a willing trial team, run a demonstration, and move on. Each effort was siloed. Multiple teams across the company were often running parallel projects on similar problems without awareness of each other.

The bigger structural flaw was timing. Innovation was typically introduced after a protocol was already finalized. Adding a new technology to an existing protocol rarely made operations easier — it added complexity without enough benefit to justify it.

There was also no systematic way to define success upfront. Teams would finish a trial and ask retrospectively whether the innovation had added value. Without predefined metrics, the answer was rarely clear.

The New Model: Problem First, Technology Second

The Global Clinical Operations Innovation Team was formed to fix this. The team is structured in two tracks. Portfolio Innovation Directors are assigned by disease area. They engage with new projects as early as nine to twelve months before the first trial is expected to start. Alongside them, three Innovation Capability Teams focus on enrollment and recruitment, patient representativeness, and differentiated trial experience — a category that includes decentralized clinical trials.

The Novartis representative described the shift as inverting the old logic. Rather than finding a technology and searching for a trial to test it on, the team now starts with the operational challenge. Capability teams conduct landscape assessments to identify vendors that might address that specific problem. Vendor qualification begins as soon as a candidate is identified — a process that typically takes six to nine months.

For a disease area like Parkinson’s, the team thinks in terms of three challenge categories: recruitment, trial flexibility, and convenience. Social media campaigns targeting patients and caregivers, patient matching databases, home nursing visits, telemedicine, and extended site hours are all evaluated as potential tools — not automatically deployed. The Novartis representative was explicit that these solutions “could” help, not that they will, and the team does not oversell technologies to trial teams.

Innovation strategy is also defined at the project level, not the trial level. For a multiple sclerosis project with three or four trials, a single strategy is developed to apply across all of them. This creates efficiency in vendor contracts and consistency in how technologies are rolled out.

Measuring What Works — and Killing What Doesn’t

The team uses a three-part measurement framework — referred to internally as “Marco’s metrics” after the colleague who designed it — that evaluates adoption, effectiveness, and impact.

Adoption tracks what proportion of eligible sites and patients accepted a given tactic. Effectiveness compares outcomes for those using the innovation against those who did not. Impact measures overall benefit against industry or external benchmarks. The Novartis representative argued that looking at any single dimension in isolation would give a misleading picture of value.

Innovations that fail to show value are stopped. The team applies this discipline deliberately — all tactics are measured, and those that cannot demonstrate return on investment are not continued. Patient matching in oncology was cited as a live example currently moving through the pipeline from pilot to potential standard practice.

At an investigator meeting for one recent trial, sites were walked through a new tactic step by step and asked whether they would adopt it. Seventy-nine percent said yes. The Novartis representative identified site buy-in as the current weak point in the model — harder to secure than patient acceptance or regulatory alignment, and the area most in need of attention going forward.

Implications for the Broader Industry

The Novartis model reflects a wider industry reckoning with the limits of pilot culture. Innovation programs at large sponsors have historically generated proof-of-concept data without pathways to scale. The result is a growing list of technologies that showed promise in one trial and then stalled.

The structural response — embedding innovation directors into planning cycles before protocol design, not after — addresses a timing problem that multiple speakers at the summit identified independently. The Novartis approach also addresses a vendor qualification bottleneck that disproportionately burdens smaller, novel vendors, who face repetitive due diligence demands from multiple sponsors asking identical questions.

Portfolio-level strategy, rather than trial-level strategy, also changes the economics. Vendor partnerships built across multiple trials in a disease area spread onboarding costs and generate richer evidence bases for evaluating whether a technology adds durable value.

What Comes Next

Eighteen months into operation, the Novartis team is supporting over 30 active projects and has innovation applied across 60 to 100 studies. All active tactics have predefined success metrics.

The Novartis representative acknowledged that demonstrating value through multiple iterations takes two to three years per tactic — meaning the full proof points for the current model are still several years away. The near-term priority is securing stronger site engagement earlier in the process. If that gap closes, the team believes its approach to scaling innovation from pilot to standard practice will become significantly more reliable.