The future of feasibility: AI-driven site selection & protocol optimisation

Clinical trial feasibility studies evaluate if a trial can be successfully conducted at a specific site, considering patient population, resources, and timelines, and typically take one month to complete. The process is often driven by fragmented systems and limited visibility into sites, investigators, and patient populations, leading to costly delays and underperforming studies.

Feasibility and site selection remain among the most significant drivers of timeline delays and budget overruns in clinical development.

But in recent months, as AI technology has evolved and improved, sponsors are increasingly turning to integrated, AI-driven approaches that are transforming feasibility from a manual, assumption-based process into a more strategic, evidence-informed model. By connecting protocol requirements with historical trial performance, real-world treatment patterns, and site and investigator data, AI can help sponsors identify higher-potential sites, improve enrolment predictability, and accelerate study startup timelines. This approach is helping reduce the feasibility process from months to weeks, allowing sponsors to get new therapeutics to market faster.

AI analysis for the right trial sites

This is particularly important as protocol complexity continues to drive recruitment challenges, costly amendments, and trial delays across clinical development. AI models can analyse historical studies, enrolment benchmarks, and operational trends to identify eligibility criteria or study design assumptions that may unnecessarily limit recruitment or create operational burden for sites. Sponsors can then use these insights to better assess enrolment expectations and protocol feasibility before significant resources are committed.

When searching for the right clinical trial sites, sponsors typically want to know:

• Which sites and investigators have proven experience in a specific therapeutic area?
• How have these investigators and sites actually performed in previous trials?
• Where can we find access to eligible, representative patients who meet our trial’s criteria?
• Which sites have the current capacity and readiness to take on a new clinical trial?

Answering these questions helps sponsors identify leading principal investigators, uncover optimal trial sites, and access competitive intel. AI can help with:

• Protocol-aware site recommendations: AI analyses protocol criteria and historical data to generate a targeted shortlist of sites already treating relevant patient populations.
• Centralised feasibility workflows: AI can help sponsors streamline the distribution, collection, and analysis of feasibility assessments, reducing reliance on spreadsheets, email chains, and manual follow-up.
• Structured, reusable insights: AI can help transform fragmented feasibility responses into standardised insights that allow sponsors to compare sites more consistently and reuse relevant findings across future studies.
• Enrolment forecasting and predictive analytics: AI models can assess historical enrolment rates, competing studies, geographic trends, and patient availability to better predict which sites are most likely to enrol successfully and stay on target.
• Representation insights: By incorporating real-world demographic and epidemiological data, AI can help sponsors identify sites with access to more representative patient populations.

One of the biggest shifts AI enables is speed to insight. Historically, feasibility assessments often required teams to manually review protocols, aggregate competitive trial information, analyse historical enrolment patterns, and coordinate across multiple vendors and internal stakeholders, a process that could take weeks. AI can now synthesise many of these inputs in minutes, helping sponsors quickly understand where eligible patients are being treated, which investigators have relevant experience, how competing studies are performing, and whether enrolment assumptions are realistic.

Improving trial cost and speed

This shift is particularly important as clinical trials become increasingly global, competitive, and operationally complex. Sponsors are often balancing accelerated development timelines with growing pressure to improve patient representation, reduce study amendments, and increase site performance. AI-driven feasibility enables organisations to make more proactive decisions earlier in development, reducing downstream inefficiencies that can significantly impact trial cost and speed.

Meanwhile, from the site perspective, this approach helps eliminate repetitive feasibility surveys and reduces time spent pursuing trials that aren’t a good fit. It also gives sites greater visibility into sponsor demand, enabling them to showcase their capabilities, highlight availability, and connect with more relevant, better-matched study opportunities.

For site staff already operating under significant administrative burden, reducing duplicative outreach and manual feasibility workflows can help preserve resources and improve engagement with sponsors. More connected, continuously updated site intelligence also allows sites to better communicate their current capabilities, patient access, and trial readiness, helping sponsors identify stronger-fit partners more efficiently.

Here's a real-world scenario: a leading biopharma organisation needed to strengthen its internal site and Principal Investigator (PI) selection process to ensure the best investigators and sites were considered for their studies. The company implemented an AI-driven feasibility and site selection workflow, accelerating site identification and benchmarking across studies. This new workflow reduced its site selection time from 12 days to 6 days - shaving about 20% off the average one-month feasibility study process.

AI and connected data are reshaping feasibility and site selection across the clinical development lifecycle, allowing sponsors to move beyond fragmented inputs and assumption-based planning towards a more predictive, evidence-driven approach that improves enrolment confidence and startup efficiency.

As adoption continues to accelerate, organisations that embrace AI-enabled feasibility and protocol optimisation will be better positioned to launch studies faster, reduce costly delays, and bring therapies to patients with greater speed and precision. The future of feasibility is no longer reactive; it is predictive, connected, and increasingly intelligent.

About the author

Ariel Katz is the co-founder and CEO of H1, a healthcare technology company focused on connecting life sciences organisations, healthcare providers, researchers, and institutions through data-driven insights. Before founding H1, Katz launched ResearchConnection, a platform designed to connect students with research opportunities, which was later acquired. Recognised for his innovation in healthcare technology, he was named to Forbes 30 Under 30 and has been featured among emerging leaders in digital health.