Closing the Gap: How AI is Solving the Supply-Demand Disconnect in Clinical Trials

The Central Problem: The Disconnect Between Trial Design and Patient Reality

A review of two pivotal industry documents from December 2025 reveals a shared, universal crisis
in clinical research: a systemic misalignment between clinical trial operations and the actual
diversity of patient populations.

According to a 2025 study on gynecological cancers (Differences in Clinical Trial Availability vs Distribution
of Disease Among Gynecological Cancers) published in CancerNetwork, there is a profound "gap in cancer
health disparities" where trial availability does not reflect the "distribution of disease" within the community.

Simultaneously, the Zurich based "OCT DACH 2025" conference report on outsourcing in Clinical Trials,
focusing on the integration of AI technology, achieving trial equity, and navigating geopolitical instability)
highlights that while the industry is eager to adopt technology, it struggles with "unforeseen
circumstances" and the urgent need for "true equity in clinical trials".

Both documents identify that current manual methods and strategic guesswork are failing to
match the right trials to the right sites and patients. The central problem is not a lack of science, but
a lack of intelligence in how trials are distributed and recruited, leading to the exclusion of minority
populations and operational inefficiencies.

Trially.ai uses Large Language Models (LLMs) to fix the disconnect between written study protocols
and the reality found in electronic health records (EHRs). The following breakdown details how Trially
removes three major roadblocks to running successful clinical trials: the supply-demand disconnect,
the equity gap, and the overwhelming burden of manual recruitment.

Challenge 1: The Supply-Demand Disconnect

The Problem: In the CancerNetwork study, authors Andrea I. Nañez, MD, and colleagues
identified that trial availability is not "equitably distributed by the ethnic and racial makeup of the
surrounding community". Specifically, while Uterine cancer had the highest incidence in the
catchment area (predominantly affecting minority women), it accounted for a small proportion of
trial enrollment, whereas Ovarian cancer (predominantly affecting White women) was
overrepresented. The research underscores that investigators effectively act as gatekeepers,
wielding significant influence over whether clinical studies are truly within reach for diverse and
marginalized populations.

The Trially Solution: Trially resolves this misalignment through its Trially Intelligence and
Pipeline Radar features.

• Feasibility based on Reality: Instead of relying on guesswork, Trially integrates directly
  with a site's Electronic Health Record (EHR) to generate "exact patient counts" for
  potential studies,. Its "Pipeline Radar" and "Feasibility Analytics" features use AI to
  analyze the specific demographics and clinical conditions of the local population—
  including unstructured physician notes—to proactively identify trials that match the
  patients the site is already treating,,. This ensures sites only pursue studies where they
  have scientifically verified candidates, eliminating the "Trial Site Mismatch" that leads to
  enrollment failures.
  
  

• Population Analytics: By analyzing a site’s specific population by "age, gender, diversity,
  meds, conditions, etc.", Trially ensures that investigators can select trials that actually
  match the disease burden of their community, correcting the disparity Nañez, MD
  identified where trial volume did not reflect clinical volume.

Challenge 2: The Urgent Need for True Equity

The Problem: During the 2025 OCT DACH summit, Pfizer’s Global Patient Engagement Lead, Maria
Rigoroso-Brandt, cautioned that despite recent industry advancements, the sector has not yet realized
the goal of "true clinical trial equity" and significant work remains. She specifically noted that data
analysis regarding demographics "should not be an afterthought anymore". Similarly, the
CancerNetwork study explicitly calls for "greater intentionality in the selection of clinical trials... to
ensure that the selected trials fit the needs of their patient population".

The Trially Solution: Trially operationalizes equity by removing the manual bias and labor
constraints that often lead to exclusion.

• Finding the "Needles in the Haystack": Trially’s matching engine is designed to "find
  needles in your haystack", identifying eligible patients from underrepresented groups that
  manual chart reviews often miss.
  
  

• Bias Reduction: By utilizing AI that parses "Any Source - EHR Included", Trially ensures
  that recruitment is based on strict clinical criteria rather than subjective selection, directly
  addressing the "recruitment barriers" and opportunity gaps cited in the gynecologic
  oncology study.

Challenge 3: Manual Workload vs. Technological Trust

The Problem: The industry is facing a crisis of efficiency. João Gonçalo Nascimento of Pfizer
noted at OCT DACH 2025 that while AI is advancing, the industry is "still trying to figure out the
best use cases" to reduce the burden on trial sites. Additionally, AstraZeneca’s Piotr Maslak
identified a significant obstacle regarding confidence in automation, stating: "Instead of looking at
the data, we’re typing in a prompt into AI and getting an output, which some people do not trust".
This challenge mirrors the operational hurdles cited in the CancerNetwork research, which argues
that "understanding why open clinical trials fail to accrue patients is important for understanding
institution-level barriers".

The Trially Solution: Trially resolves the tension between manual burnout and AI skepticism
through high-precision transparency.

• Transparency and Trust: To dismantle the "black box" skepticism noted by
  AstraZeneca’s Piotr Maslak—who warned that users distrust AI when they "type in a
  prompt... and get an output" without seeing the data—Trially eliminates the need for blind
  faith through "Transparent Match Detail." Instead of providing a static list of candidates,
  Trially’s engine enables users to "Triple Click" directly into specific Inclusion/Exclusion
  criteria, instantly revealing a "synthesis of physician notes" and unstructured text that
  justifies the recommendation. By exposing the "why" behind every match—proof that the
  AI has accurately parsed complex medical history—Trially allows investigators to validate
  findings in real-time, bridging the gap between automated efficiency and clinical trust.
  
  

• Operational Efficiency: Trially is proven to drive a "90% reduction in EHR chart review
  hours". This directly alleviates the burden on investigators, allowing them to focus on
  engaging diverse populations rather than drowning in paperwork—a key factor in resolving
  the enrollment failures cited in both source documents.

Conclusion

The data from late 2025 is clear: the clinical trial landscape is struggling to align scientific
opportunity with patient reality. Whether it is the "significant disparities" in gynecologic cancer
enrollment identified by CancerNetwork or the geopolitical and equity pressures discussed at OCT
DACH 2025, the industry requires a structural shift.

Trially.ai serves as this structural shift. By moving from manual guesswork to AI-driven "Pipeline
Radar" and "Feasibility Analytics", Trially ensures trials are not just available, but are effectively
matched to the diversity of the patient population. It answers the call from experts like Nañez and
Rigoroso-Brandt for greater intentionality and data-driven equity, proving that technology is the
key to closing the gap between the trials we design and the patients we serve.