Outcomes & Evidence

Enrollment Impact

What changes when CRO feasibility teams use structured EHR matching instead of manual chart review — directional outcomes from pilot programs.

Key Metrics

What changes with Cohortbridge

These directional outcomes reflect results from internal pilot programs with participating health system and CRO partners. They are not peer-reviewed benchmarks.

Days
not weeks
Initial cohort identification
From EHR matching run to first cohort estimate
70%+
Chart review redirected
From discovery to eligibility confirmation
Earlier
Sponsor enrollment estimates
Prospective cohort data before site activation
Chart Review

From discovery to confirmation

Manual chart review has two distinct phases: discovery (which patients might be eligible?) and confirmation (are these patients actually eligible based on full chart review?). In the traditional model, CRAs spend most of their time on discovery — reviewing charts to find candidates that should have already been identified from structured EHR data.

Cohortbridge handles the discovery phase. CRA chart review shifts to confirmation — a smaller, higher-value activity focused on verifying eligibility for patients the matching engine has already identified as likely candidates.

Before and after comparison chart showing enrollment timeline reduction with Cohortbridge patient matching

Without Cohortbridge

Chart discovery
~70%
Eligibility confirm.
~30%
Timeline
6–8 wk

With Cohortbridge

Chart discovery
~20%
Eligibility confirm.
~80%
Timeline
Days

Directional ranges. Results depend on protocol complexity, EHR data completeness, and site characteristics.

Case Studies

How CROs have used Cohortbridge

These case summaries describe how CRO feasibility teams have applied Cohortbridge to specific trial types. Details are representative of pilot program experiences.

Cardiometabolic — Phase II

Morrow Clinical Group

A 12-site cardiometabolic Phase II study with narrow inclusion criteria (HbA1c 7.5–10%, no prior insulin, eGFR ≥ 45). Initial site estimates based on historical enrollment predicted 6–8 eligible patients per site.

After running Cohortbridge against Epic populations at 8 connected sites, the team identified that 4 sites had significantly fewer eligible patients than estimated (2–3 each) and 2 sites had substantially more (12–15 each). Site allocation was adjusted before activation — avoiding 4 under-performing site activations and adding 1 higher-yield site.

Site selection adjusted pre-activation
Rare Disease Oncology — Phase III

Halcyon Research Partners

A rare disease oncology Phase III study with an estimated prevalence of 1–2 patients per 10,000 in the general population. Conventional feasibility approaches estimated 1–2 eligible patients per site per year at 15 sites.

Cohortbridge matching across 4 health system EHR networks (combined ~2.3M records) identified a concentrated population at 2 academic medical centers with rare disease specialty practices — sites not on the original activation list. Protocol-specific rare diagnosis codes surfaced patient clusters not visible in historical enrollment data.

Previously unidentified high-yield sites found
Methodology Note

What these metrics represent

The enrollment impact data presented on this page reflects directional outcomes from Cohortbridge's internal pilot programs with participating health system and CRO partners in 2024–2025. These outcomes are:

  • Not peer-reviewed or independently verified
  • Not representative guarantees — individual results vary by protocol complexity, EHR data completeness, and site characteristics
  • Reported as directional ranges, not point estimates
  • Based on a limited sample of pilot programs — not a large-scale randomized trial

Cohortbridge is a seed-stage company. We believe the directional evidence from our pilot programs is compelling, but we report it honestly as pilot data — not as established clinical research benchmarks.

See a cohort analysis for your protocol

We'll run a sample analysis showing how Cohortbridge would approach your protocol's inclusion/exclusion criteria against a connected EHR dataset.