Published on 15/11/2025
Protocol Feasibility Lessons That Turn Design Risk into Operational Reliability
Feasibility Reframed: From “Can We?” to “Will It Work Reliably Across Countries and Sites?”
Feasibility is not a one-time survey or a quick recruitment guesstimate; it is a cross-functional stress test of whether a proposed protocol can be executed reliably by real participants, real investigators, real systems, and real vendors under regulatory scrutiny. The quality objective is simple to state and hard to deliver: design a study that protects participants, generates credible endpoints, and does so predictably at scale. A feasibility
Regulatory anchors that shape feasibility choices. A defensible feasibility process uses internationally recognized principles. The proportionate, risk-based orientation of the International Council for Harmonisation (ICH) guidance reminds teams to concentrate controls on critical-to-quality factors such as informed consent, endpoint timing and standardization, safety reporting clocks, and data integrity. In the United States, expectations around protocol adherence, investigator responsibilities, consent, and trustworthy electronic records/signatures are captured throughout FDA clinical trial protection resources. For European programs, feasibility must also anticipate the operational cadence under EMA clinical trial guidance (e.g., what will constitute a “serious breach,” how safety and data reliability are demonstrated). Ethics considerations—participant burden, comprehension, privacy, and fair risk–benefit—should be visible in every feasibility decision, consistent with WHO research ethics guidance. For Asia-Pacific delivery models, align terminology and expectations with PMDA clinical information and TGA clinical trial guidance so country differences do not become late surprises.
Where feasibility most often fails. Post-mortems show recurring causes: endpoints that require narrow windows or specialized equipment unavailable at many sites; long, complex visits that exceed clinic capacity; eligibility criteria that are “medically tidy” but rare in practice; decentralized elements added without verifying identity, privacy, and device reliability; dependency on a single vendor for a CtQ activity; and data interfaces (EDC, eCOA, IRT, imaging, safety) designed without a reconciliation plan. Each failure mode converts into protocol deviations later—missed windows, consent errors, device outages, temperature excursions, or unblinding incidents. Feasibility success means eliminating or mitigating these risks before first patient in.
Lessons that reframe the conversation. First, flip the order: do not write the schedule of assessments until after you simulate a day-in-the-life for participants and sites. Second, model variability (not averages): travel, clinic capacity, diagnostic lead times, courier performance, and power outages all fluctuate. Third, choose “robust by design” over “perfect on paper”—allow flex where it does not harm inference, and add precise rules where integrity would otherwise drift. Fourth, treat feasibility evidence as inspection-facing artifacts, not slideware. If an inspector asks, “Why is this window ±5 days?” your team should produce the modeling memo and risk rationale within minutes.
The feasibility bill of materials. A complete package covers: country and site availability of equipment and specialists; participant journey and burden analysis; visit schedule stress testing; decentralized/remote readiness; safety reporting logistics; IP supply and temperature control; data lineage and interface ownership; privacy and redaction practices for remote review; and quality measures (QTLs and KRIs) wired into the monitoring plan. When these components are present, amendments decline, recruitment stays on plan, and deviations trend flat rather than exponential.
Designing the Study Around Reality: Patient, Site, Data, and Supply Feasibility
Patient journey and burden. Map the path from screening to last visit with real travel times, work/family constraints, and common comorbidities. Convert vague goals (“reduce burden”) into measurable targets: average on-site time per visit, total blood volume, number of questionnaires, night-time alarms from wearables, and the number of invasive procedures. Where burden is high, apply substitutions that preserve inference: fewer but better standardized assessments; home health for non-critical draws; asynchronous diaries rather than rigid daily windows; and travel support calibrated to local costs. Feasibility improves when participants can say “yes” without heroic scheduling.
Site capacity and workflow. A protocol may be scientifically elegant yet operationally impossible in a Tuesday clinic. Stress test: can a coordinator complete consent with comprehension checks, prep the room, perform endpoint measures, package specimens, and close queries while the PI performs safety evaluations and the pharmacist reconciles IP? If the answer is “only with overtime,” redesign. Combine procedures into one room, shorten instrument warm-up, enable pre-visit remote PROs, or stagger windows. Write time-and-motion estimates directly into the feasibility memo and use them to drive the schedule of assessments.
Endpoint operationalization. Measurement integrity is non-negotiable. Confirm instrument availability, calibration requirements, rater qualification pathways, and backup plans for outages. If imaging is critical, assess reader network capacity, acquisition parameters, and courier timeliness for hard media (where used). For eCOA/wearables, specify firmware control, battery/charging cadence, and a device swap process. Create “validity flags” in the data specification (correct instrument version, standardized conditions met) so downstream analyses can apply rules mechanically.
Decentralized and hybrid elements. Remote workflows succeed when identity, privacy, and logistics are rehearsed. Validate eConsent identity proofing and language support; script tele-visit privacy checks; document device activation and data sync steps; and define direct-to-patient (DtP) shipment controls with temperature loggers and photographable chain-of-custody. Choose which assessments truly require on-site presence. State in the protocol where substitution is allowed and how to document it in source—clarity here prevents deviations later.
Supply chain and pharmacy. Model inventory, expiry, and temperature risk under realistic delivery variability. Link IRT strategies to site enrollment velocity and visit cadence; simulate weekend/holiday impacts; and pre-authorize emergency resupply routes. For products sensitive to excursions, publish a simple decision tree for quarantine or release, and rehearse it at site initiation. Fail feasibility here and you will see dosing delays, accountability mismatches, and unplanned deviations.
Data flow and interfaces. Draw a data lineage map: origin (source), capture system, transformations, and reconciliation frequency. Assign an “interface owner” for each connection (EDC↔eCOA, EDC↔IRT, safety↔EDC, imaging↔EDC). Define how exceptions surface (dashboards, tickets) and who resolves them. Require audit trails, signature manifestation, time synchronization, and export formats that your statistics and QA teams can verify quickly. Feasibility does not end with recruitment; it includes the practicality of proving provenance months later.
From Assessment to Design: Translating Feasibility into Protocol Language and Start-Up Plans
Write feasibility into the protocol, not just the playbook. Convert feasibility findings into explicit text: windows with rationale; substitution rules for decentralized workflows; identity and privacy steps for remote activities; calibration and rater requirements; courier/temperature expectations; and data reconciliation responsibilities. Where flexibility is acceptable, say so; where rigidity protects endpoint integrity, make it unmistakable. Ambiguity is the fastest path from “feasible” to “deviation.”
Quality tolerance limits (QTLs) and key risk indicators (KRIs). Define a small set of QTLs tied to CtQ risks (e.g., primary endpoint window misses <1%; median hours from SAE awareness to initial submission below a defined limit; eligibility adjudication errors <Y per 100 screenings). At site level, track KRIs such as consent errors per 50 consents, eCOA missingness in a rolling window, device firmware changes without validation, and reconciliation exceptions older than seven days. Publish thresholds and owners in the monitoring plan so the feasibility logic drives oversight behavior.
Country and site strategy. Use feasibility evidence to prioritize countries with available equipment, predictable ethics timelines, and lower logistics volatility. Within each country, select sites that can demonstrate prior performance on analogous endpoints and technologies. Require a “show me” drill during qualification: retrieve consent examples, demonstrate endpoint equipment, produce eCOA dashboards, and walk through IRT/temperature exception handling. Favor sites that can produce evidence quickly; that speed foretells monitoring success.
Start-up and training. Turn feasibility into onboarding materials: job aids for consent, eligibility, endpoint procedures, safety clocks, and decentralized privacy; micro-modules timed before first recruitment; and simulations that rehearse the hardest steps (eConsent identity, imaging acquisition, device swap, IP excursion). Gate Delegation of Duties and system access on completion and observed competence. Document training language on certificates and provide bandwidth-light versions for constrained regions.
Vendors and governance. Flow feasibility assumptions into quality agreements and SOWs: exportable evidence with audit trails; time-boxed support SLAs; release gates for firmware/app updates; and participation in retrieval drills. Establish a cadence—weekly site/CRO huddles on readiness, monthly study reviews on QTL/KRI status, and quarterly cross-study steering to retire vanity metrics and update exemplars. When an amendment or technology release shifts risk, update job aids and micro-modules and record “what changed and why.”
Budget and time realism. Good feasibility frequently adds cost (extra calibration, wider windows, courier redundancy) while reducing the far larger costs of rescue amendments and delays. Make the trade visible: show the expected reduction in deviations, queries, and rework minutes; quantify courier risk avoided; and display an “amendment savings” estimate. Executives reward designs that avoid late surprises and protect the probability of success.
Metrics, Pitfalls, and a Ready-to-Use Feasibility Checklist
What to measure to prove feasibility worked. Track leading indicators rather than vanity counts. Useful KPIs include: percentage of sites that passed a retrieval drill before first patient in; percentage of decentralized sessions with documented identity and privacy checks; proportion of endpoint assessments meeting standardized conditions on first attempt; median hours to initial SAE submission; eCOA device uptime; reconciliation exceptions closed within SLA; and time-to-green after a red KRI at a site. Also trend amendment frequency and the share of deviations that map to feasibility assumptions—you want both falling.
Common pitfalls and the countermeasures that work. (1) Design by average: Plans assume mean visit length or typical courier times; counter with variability modeling and buffers where integrity allows. (2) Over-tight windows: Beautiful on paper, brittle in practice; widen where the estimand is robust, and provide rescue assessments. (3) Single-vendor fragility: Add parallel capacity or at least emergency fallbacks. (4) Unclear decentralized rules: Write identity, privacy, and device steps into the protocol and source templates. (5) Interface ambiguity: Assign owners, frequencies, and exception handling before enrollment. (6) Training ≠ competence: Gate delegation and access on observed performance, not attendance. (7) Evidence afterthought: Treat feasibility outputs as inspection artifacts; pre-map Trial Master File locations and rehearse retrieval.
“Lessons learned” you can paste into your playbook. Start the schedule of assessments only after a day-in-the-life simulation; involve pharmacy and couriers early; write substitution and rescue rules explicitly; define QTLs/KRIs from the feasibility model; require a vendor “connection control pack” for each interface; run pre-go-live monitoring simulations; and keep a country addendum with ethics timelines and logistics variability. Above all, make the reasoning transparent: every rigid rule should protect safety or endpoint integrity; every flexible rule should be justified by equal or better reliability.
Feasibility checklist for immediate use.
- Document a participant journey with time-and-motion estimates; set burden targets (on-site minutes, questionnaires, blood volume).
- Simulate site workflows (consent, endpoint, safety, pharmacy); confirm capacity with realistic Tuesday throughput.
- Validate decentralized/remote steps: eConsent identity and language; tele-visit privacy; device activation/sync; DtP chain-of-custody and temperature logging.
- Confirm endpoint integrity: instrument availability; calibration and rater pathways; imaging parameters and reader capacity; device firmware control and swap process.
- Model supply risk: IRT strategies; shipment variability; emergency resupply; excursion decision trees and rehearsal.
- Draw data lineage: interfaces, owners, reconciliation frequency, exception surfacing; verify audit trails and time synchronization.
- Write feasibility into protocol text: windows with rationale; substitution/rescue rules; identity/privacy steps; calibration and reconciliation responsibilities.
- Set QTLs and KRIs with owners and thresholds; embed in the monitoring plan; plan dashboards with drill-downs to evidence.
- Prepare onboarding: role-specific job aids, micro-modules, and simulations; gate delegation and access on competence; localize materials.
- Pre-map TMF/ISF locations for feasibility artifacts; run a retrieval drill before first patient in; record outcomes and actions.
The inspection story. When an inspector asks, “Why is your design reliable in the real world?”, you should be able to show the simulation memos, variability models, substitution rules, identity/privacy steps, interface ownership, and the QTL/KRI wiring that links design to oversight. That narrative—grounded in internationally recognized quality and ethics principles and tuned for country-specific realities—turns feasibility from a slide into a system.