TrialKey

Study Design and Protocol Optimisation

Clinical trial feasibility—and probability of success—tested before execution.

Estimate whether the proposed design can meet its primary endpoint, then pressure-test recruitment, protocol burden, countries, sites, timelines and costs before the plan becomes an expensive commitment.

See how feasible your proposed study really is.

Bring your indication, treatment type and draft protocol or study plan, if available.

Request a trial review
See probability and feasibility together
570,000+ trials analysedEnrolment and protocol benchmarksExpert supervised
EndpointsEligibilityEnrolmentSitesCountriesVisitsTimelineBudget

Study inputs

Draft study plan or indication + treatment type

Add the protocol, synopsis, assumptions and constraints that are available today.

Hybrid ML + feasibility agent

Historical evidence guides the agent as it tests the study's connected assumptions.

Predict and stress-test

Estimate P(success) and test recruitment, burden, footprint, duration and cost together.

Human checkpoint

Expert review

Challenge the evidence, refine assumptions and rule out impractical options.

Decision-ready output

Feasible alternatives with modelled impact

Compare predicted success, feasibility risks and practical next steps.

Real-world constraints

Budget · Timeline · Regulatory requirements · Geography · Patient burden

Probability of success + feasibility

Estimate the probability of success—and see what could improve it.

TrialKey estimates the proposed design, explains the model drivers, generates alternatives and tests whether the stronger candidates can be delivered in the real world. The recommendation balances predicted success with historical evidence, study priorities and feasibility constraints.

Example study-plan optimisation

Study plan comparison

TrialKey compares every alternative with the same reference plan, estimates its probability of meeting the efficacy objective and checks whether the changes are supported by historical trial patterns.

The recommended plan balances modelled probability of success, evidence alignment, study priorities and feasibility constraints—it is not selected on probability alone.

Reference predicted success 25.9%
Probability leader 47.6%+21.7 pp
Recommended after evidence weighting 47.3%

Candidate progression

Predicted probability of success

Recommended after full assessment
55%
45%
35%
25%
15%
ReferenceCandidate 4Candidate 8Candidate 12RecommendedLeader

Optimisation progress

  1. Starting plan assessed
  2. Alternatives prepared
  3. Alternatives compared
  4. Recommended plan selected

Plan comparison

  1. 1Probability leader47.6%
  2. 2Recommended balanceSelected47.3%
  3. 3Alternative candidate47.1%
  4. 4Evidence-aligned option39.1%
  5. 5Conservative option39.0%

Eighteen alternative study plans assessed. Historical evidence is used as a separate grounding check alongside the modelled probability of meeting the study objective.

Recommended only after feasibility and expert review

What changed

Plain-language decision readout

Evidence check passedRecommended uplift +21.4 pp

Generated design changes were compared with 1,200 historical eligibility criteria before the recommended plan was surfaced.

What looks better

The recommended candidate increases the modelled probability of meeting the efficacy objective by 21.4 percentage points while retaining the core disease-confirmation and outcome anchors.

What to check

Confirm that the refined eligibility criteria remain clinically appropriate and do not create a new patient-access or recruitment constraint.

Recommended next step

Review the changed eligibility lines with the clinical and operational subject-matter experts before adopting the candidate as the study-plan recommendation.

Eligibility comparison

Reference plan versus recommended plan

Illustrative ALS study example

Reference plan

Eligibility assumptions

  • Broad eligibility wording
  • Respiratory criteria not consistently bounded
  • Overlapping treatment washout requirements
Current enrolmentCurrent footprintCurrent endpoints

Recommended plan

Clarified eligibility

  • Age 18–75 years at screening
  • Defined ALS diagnostic criteria
  • SVC between 42% and 90% predicted
  • Explicit ALSFRS-R screening threshold
  • Harmonised treatment washout rules
Endpoint retainedEligibility refinedRecruitment re-tested

Clinical promise and operational feasibility are not the same thing.

A study can have a promising modelled probability of meeting its endpoint but still be difficult to recruit or deliver. A highly recruitable study can still have weak statistical or clinical prospects. TrialKey keeps both questions visible and ranks alternatives across them.

Estimate probability of success

Model the probability that the proposed design will meet its primary efficacy endpoint, with a confidence range rather than a single unsupported score.

Identify the model drivers

See which study-design variables, evidence signals and assumptions contribute most to the predicted outcome.

Recommend feasible alternatives

Generate and compare design options that may improve predicted success without creating an impractical protocol or operating plan.

Quantify the trade-offs

Compare probability, recruitment, patient burden, timeline, cost and geographic footprint before selecting a preferred scenario.

Clinical trial feasibility assessment

Feasibility is a connected system, not a site questionnaire.

Clinical trial feasibility depends on the interaction between protocol design, patient availability, recruitment assumptions, site capability, country selection, timelines and budget. TrialKey evaluates those decisions together so one optimistic assumption does not undermine the entire study plan.

Patient and eligibility feasibility

Test whether the proposed population and eligibility criteria leave a credible pool of recruitable participants.

Enrolment feasibility

Benchmark recruitment rates, screen-failure assumptions and enrolment duration against relevant completed and competing studies.

Protocol feasibility

Identify visits, assessments, endpoints and operational requirements that may create avoidable patient or site burden.

Country feasibility

Compare patient access, investigators, competing trials, regulatory conditions and operating complexity across countries.

Site feasibility

Evaluate whether the proposed footprint can support activation, recruitment, data quality and delivery expectations.

Timeline and budget feasibility

Pressure-test whether start-up, recruitment, treatment and follow-up assumptions fit the programme's time and capital constraints.

Clinical trial delay evidence

Protocol choices compound into delivery delay.

Historical trial outcomes show that delayed completion is common—and that delay rises sharply as protocols accumulate amendments. TrialKey makes those risks visible while study alternatives can still be tested.

3 in 4finish late

Three out of four clinical trials finish late.

Completion delay is measured by comparing the date a trial registered as its planned completion with the date it actually finished.

9.8months

median completion overrun

45%

finish more than a year late

29%

finish more than two years late

Source: TrialKey internal analysis of registered clinical trial completion dates and protocol amendment histories.

Amendment burden

More protocol amendments are associated with much longer delays.

27 months late

Median delay for trials with 11 or more amendments, versus about 4 months for trials with one or two.

Approximately 6.7× longer

Protocol amendments

Median delay

1–2 amendments

4.0 mo

3–5 amendments

7.4 mo

6–10 amendments

14.1 mo

11+ amendments

26.6 mo

Historical association does not establish that amendments alone caused the delay. The pattern is used as evidence to strengthen feasibility review and protocol planning.

How it works

Evidence-led study feasibility, with expert supervision.

TrialKey combines historical clinical-trial evidence, protocol features, competing-study activity and operational assumptions to identify where the proposed plan is credible—and where it needs refinement.

01

Structure the proposed study

Start with a protocol, synopsis or draft study plan—or with the indication, treatment type and development concept when the design is still taking shape.

02

Benchmark the critical assumptions

Compare eligibility, enrolment, sites, countries, visits, duration and other feasibility drivers with relevant historical and competing trials.

03

Predict success and model feasible scenarios

Estimate the proposed design's probability of success, then test how changes to the population, endpoints, burden, footprint and timeline affect both clinical prospects and execution risk.

04

Recommend and review alternatives

The system compares evidence-supported options, then clinical and operational experts challenge the results and convert the strongest feasible scenario into recommendations a team can defend.

From protocol to operations

Find the assumptions most likely to delay the study.

A credible feasibility plan separates evidence-supported assumptions from numbers that are merely convenient for the business case. TrialKey shows where the proposed protocol and operating model are most exposed.

Protocol and patient burden

Assess whether eligibility criteria, visit schedules, procedures, endpoints and follow-up requirements make participation or delivery unnecessarily difficult.

Recruitment and timeline assumptions

Compare recruitment rates, screen failures, activation timelines and study duration with evidence from relevant trials rather than relying on a single-point forecast.

Country and site footprint

Identify where patient access, investigators, competing studies and operational capability are most likely to support high-quality enrolment.

Feasibility outputs

Replace a yes-or-no answer with a decision-ready plan.

Clinical trial feasibility should explain why the plan is credible, where uncertainty remains and what can be changed. TrialKey produces a traceable assessment that clinical, operational, regulatory, finance and governance teams can review.

Evidence rangesRisk driversScenario comparisonsExpert recommendations

What you receive

  • Modelled probability of success, confidence range and key drivers
  • Clinical trial feasibility assessment and executive scorecard
  • Protocol burden and operational-risk review
  • Evidence-based enrolment and screen-failure assumptions
  • Country and site-footprint recommendations
  • Competing-trial and patient-access analysis
  • Alternative feasibility scenarios and trade-offs
  • Expert-reviewed recommendations for the study plan

Where feasibility fits

The bridge from study design to country and site selection.

TrialKey uses feasibility to turn a proposed design into realistic operating assumptions, then carries those assumptions into country and site selection.

Feasibility layer

Clinical Trial Feasibility

Test whether the design can recruit and operate within real constraints.

Questions

Clinical trial feasibility FAQs

What is clinical trial feasibility?+

Clinical trial feasibility is the assessment of whether a proposed study can recruit the required participants, operate across its intended countries and sites, meet its timeline and remain deliverable within clinical, regulatory and budget constraints.

What is a study feasibility assessment?+

A study feasibility assessment tests the assumptions behind a clinical study before execution. It can examine patient availability, eligibility criteria, recruitment rates, protocol burden, country and site capacity, competing trials, timelines and costs.

How is protocol feasibility different from site feasibility?+

Protocol feasibility asks whether the study design itself is practical for patients, investigators and sites. Site feasibility asks whether specific research sites have the participants, capabilities, capacity and operating conditions needed to deliver that protocol.

Can TrialKey assess feasibility before a protocol is written?+

Yes. TrialKey can begin with an indication, treatment type, development phase and draft study concept. A protocol, synopsis or study plan can be incorporated when available, but it is not required to begin an early feasibility assessment.

How does AI support clinical trial feasibility?+

AI and machine learning can compare the proposed study with historical and competing trials, identify influential feasibility drivers and test alternative assumptions. TrialKey combines this analysis with expert review rather than relying on an automated score alone.

How does probability of success relate to clinical trial feasibility?+

They answer connected but different questions. Probability of success estimates how likely a study design is to meet its primary efficacy endpoint. Clinical trial feasibility tests whether that design can recruit, operate and finish within real-world constraints. TrialKey evaluates both so a clinically promising design is not recommended if it is impractical to deliver.

Does a feasibility assessment guarantee recruitment performance?+

No feasibility assessment can guarantee recruitment or delivery. TrialKey provides evidence-based ranges, identifies material risks and compares alternatives so teams can make more defensible planning decisions before committing resources.

Why do protocol amendments matter to clinical trial feasibility?+

Higher protocol-amendment counts are associated with longer clinical-trial completion delays in TrialKey's readiness analysis. Amendments can add review cycles, site retraining, document and system changes, resupply work and recruitment disruption. TrialKey uses historical patterns as evidence—not as proof of causation—to identify design pressure points and compare alternatives before execution.

Evaluate study feasibility

Test the plan before start-up begins.

Bring us your indication, treatment and proposed study assumptions. We will show how TrialKey can structure the evidence, expose the material feasibility risks and compare practical alternatives.