Published on 15/11/2025
Greenlight Checklists and Go/No-Go Decisions—From Readiness Evidence to Safe First-Patient-In
Purpose, Principles, and the Global Frame for Greenlight Decisions
Greenlight is not a date—it is a documented threshold of evidence that a site can protect participants and preserve the integrity of critical endpoints from day one. A go/no-go decision translates that threshold into action: begin recruitment now, pause until specific risks are controlled, or restage the activation sequence. This is where start-up governance proves its value. A strong greenlight system is concise, auditable, and proportionate to risk; it replaces informal reassurance
Why proportionate control matters. Greenlight criteria should reflect a quality-by-design posture in line with the spirit of harmonized expectations discussed by the International Council for Harmonisation. In the U.S., teams often calibrate investigator responsibilities, consent logistics, and trustworthy records against educational materials hosted within the FDA’s clinical trial protection resources. For Europe and the UK, authorization cadence and public-facing duties shape readiness elements that must be documented; planning is commonly informed by information made available by the European Medicines Agency. Ethical touchstones—respect, voluntariness, and confidentiality—are reinforced by practical perspectives accessible through the World Health Organization’s research ethics guidance. For Japan and Australia, ensure readiness language and artifacts remain coherent with orientation material provided by the PMDA and Australia’s Therapeutic Goods Administration.
ALCOA++ and inspection posture. Regulators and auditors ask three questions at activation: (1) Is the site prepared to protect participants and the blind today? (2) Is the evidence traceable from dashboard to artifact with ALCOA++ attributes—attributable, legible, contemporaneous, original, accurate, complete, consistent, enduring, and available? (3) If something goes wrong, can the sponsor demonstrate that foreseeable risks were recognized and controlled? A regulator-ready greenlight system answers yes by making each criterion binary (met/not met), evidence-backed, and mapped to named owners.
What a greenlight checklist must decide. A credible checklist goes beyond “slides were presented.” It confirms that people are competent and authorized, places are equipped and validated, products are controlled and tracked, and processes are synchronized—including decentralized workflows and device/diagnostic specifics. It also encodes firewalls (blinding/privacy), first-day rehearsals, and early-ramp indicators so go-live is both safe and sustainable.
Meaning of approval—make signatures informative. Every sign-off should capture what was approved, not only who signed: “clinical accuracy verified,” “logistics validated,” “UAT completed (positive and negative tests),” “ALCOA++ check passed,” “privacy and identity safeguards configured.” When signatures carry meaning, accountability is explicit and inspection-ready; when they do not, reviewers are left to infer what a signature meant.
System, not heroics. Replace ad-hoc spreadsheets with a controlled template and a short evidence pack per site. Use deterministic file names and required metadata (country, site, process, version, effective dates, owner). Wire the checklist to your start-up dashboard so each status tile clicks through to a single artifact in the eTMF/ISF. Practice a five-minute retrieval drill before greenlight—if your team cannot produce the chain immediately, fix filing and metadata now, not during inspection.
Building the Greenlight Checklist—People, Places, Products, Processes, and Firewalls
People—competence and authority. Confirm investigator CV/license currency (within the activation window), GCP training attestations, financial disclosures, and a reconciled delegation of duties log with start/stop dates. Validate that system access (EDC, IWRS/IRT, imaging, eConsent, safety portals) matches roles. Require short knowledge checks for endpoint-critical steps (eligibility windows, primary assessment timing, unblinding requests). Evidence: CV/license PDFs, GCP certificates, signed delegation log, role-based access report, quiz results with thresholds (e.g., ≥80% overall and 100% on critical items).
Places—equipment, logistics, and environmental controls. Pharmacy readiness (temperature mapping, alarm thresholds verified, excursion drill executed), investigational product receipt and quarantine/release rules, waste and returns plans. Imaging/lab readiness (test file upload accepted by the reading/central lab; reference ranges loaded; back-up slot plan). Courier accounts live, dry-ice/hazardous-goods constraints documented, and first pickup tested. For hybrid or decentralized workflows, confirm mobile nursing availability, safe kit storage, return logistics, and help-desk scripts. Evidence: temperature maps and alarm screenshots, test upload receipts, courier bill of lading, hazard declarations, mail-back instructions.
Products—control, labeling, and configuration. Verify product/device labeling, kit assortment, and expiry windows; document stock reconciliation between physical inventory and IWRS/IRT. For device or diagnostic studies, confirm hardware model and firmware/software versions, site acceptance tests, disposables/consumables inventory, and quarantine/release rules if configuration issues arise. Evidence: label proofs, packing lists, stock reports, acceptance test logs, and a version table stored with the site’s configuration sheet.
Processes—protocol, consent, privacy, and safety. Ethics/authority approvals on file; informed consent version in effect (threaded to approval dates) and localized as required; recruitment materials approved; privacy and cross-border data transfer narratives in place; SAE clock rehearsal completed; emergency unblinding request path tested; eConsent identity verification demonstrated (both in-person and tele-consent if used). Evidence: approval letters, localized consent pack manifest, readability/translation certificates where applicable, privacy statement, unblinding test record, identity-verification screenshots.
Firewalls—blinding and access segregation. Separate unblinded pharmacy/IWRS roles from blinded coordinators and raters; restrict read access to code-break logs and unblinded stock; document a clean hand-off for temperature excursions and resupply decisions. Evidence: role matrix, access approvals, and a one-page firewall description filed with the SIV minutes.
First-day rehearsals—prove the path works. Conduct a table-top walk-through or simulation of consent → eligibility confirmation → randomization → first dose/first procedure, including rescue paths for common errors (e.g., visit outside window, scanner downtime). Include decentralized steps—tele-visit connectivity check, mobile nurse scheduling lead-time, identity re-verification. Evidence: SIV minutes with outcomes, mock run sheets, screen captures, and a short defect list with owners and due dates.
Risk-based acceptance—binary thresholds and conditional activation. Define objective pass criteria (e.g., “training completion ≥95% with no critical gaps,” “EDC and IWRS UAT sign-offs filed with at least one negative test each,” “courier test pickup completed,” “identity verification success rate ≥98% across two scenarios”). If some non-critical evidence is pending, allow conditional activation with time-boxed items and an explicit pause rule if milestones slip. Record every exception with rationale and owner so inspectors see proportionate control, not improvisation.
Five-minute retrieval drill—make it muscle memory. Before signing the greenlight memo, pick one dashboard date and retrieve: delegate log, user provisioning report, UAT logs, pharmacy mapping/alarms, imaging test receipt, courier test proof, localized consent in effect, ethics approval letter, privacy/identity evidence, and the signed go/no-go decision. If retrieval exceeds five minutes, correct metadata, refactor filing locations, or enforce a single record-of-record to eliminate duplicates.
Go/No-Go Governance—Decision Meetings, KRIs, QTLs, and Escalation
Small, named board with the meaning of approval. Keep the decision group focused: Site Operations Lead (chair), Regulatory/Ethics Lead, Pharmacy/Supply Lead, Data Systems Lead (EDC, IWRS/IRT, eConsent), and Quality. Each signature must include a short phrase stating what the approver reviewed and approved. The meeting runs on evidence, not updates: the checklist and artifact links are screened beforehand; the board adjudicates exceptions and signs the memo.
Agenda that ends with a decision. (1) Confirm binary criteria met; (2) Review exceptions and conditional items with owners and due dates; (3) Confirm firewalls and privacy controls; (4) Review early-ramp readiness (imaging blocks, courier capacity, navigator staffing, interpreter coverage); (5) Decide: go (release IP), conditional go (time-boxed items, pause rule), or no-go (specific defects to correct). Decisions are recorded in a one-page memo and filed to the eTMF/ISF with the evidence pack.
KRIs that warn before go/no-go fails. Define leading indicators and thresholds that automatically open actions: (1) Delegation vs. access drift (mismatch rate >0% on reconciliation); (2) Consent version drift (localized version not threaded to approval dates); (3) UAT gaps (no negative tests recorded); (4) Pharmacy exceptions (mapping or alarm verification pending); (5) Imaging readiness (test upload not accepted); (6) Courier exceptions (failed test pickup or missing hazardous-goods documentation); (7) Identity-verification failures (>2% failed attempts in pilot). Red KRIs trigger either rapid containment (24 hours) or a no-go recommendation unless closed.
Quality Tolerance Limits (QTLs) tied to early ramp. Convert the most consequential KRIs into QTLs: no consent within 21 days of activation; ≥10% missed endpoint windows among first 10 participants; ≥15% courier exceptions over two weeks; or eConsent restarts >10% for tele-visits. Crossing a QTL forces the board to meet and either contain (local fix), correct (design/process changes), or communicate (resequence activation, add resources). Record the decision with rationale and link it to the triggering tile.
Conditional activation with explicit pause rules. If non-critical items remain, allow go-live with hardholds, for example “activate but do not randomize until pharmacy alarm verification filed,” or “consent only; randomize after imaging QC passes.” Set hard dates—if unmet, the site pauses automatically. Conditional logic is proportionate, transparent, and self-enforcing.
Vendor oversight. If readiness relies on vendors (EDC/eConsent/IWRS, imaging cores, couriers, home health), encode obligations in SOWs: immutable logs, weekly evidence feeds, synchronized clocks, and participation in retrieval drills. Require at-risk fees if persistent red metrics delay go/no-go decisions.
Change control and communication. Treat the checklist and thresholds like controlled code. Version changes, publish a “what changed and why” memo, retrain impacted roles, and update the metric dictionary. After each decision, communicate a short “greenlight bulletin” to the site: evidence reviewed, items still pending, pause rules, first-day reminders, and help-desk contacts.
Operating Playbook—Templates, Pitfalls, 30–60–90 Plan, and a Ready-to-Use Checklist
Templates to standardize speed. Provide (a) the greenlight checklist with binary criteria; (b) a one-page go/no-go memo template that captures decision, rationale, exceptions, owners, and dates; (c) a firewall description template showing blinded/unblinded roles and system access; (d) a first-day run sheet for consent→eligibility→randomization→first dose/procedure; and (e) a retrieval drill script with a 5-minute timebox and “fix filing now” instruction if the drill fails.
Common pitfalls—and durable fixes.
- “We presented it” instead of “they can do it.” Fix with outcome tests (dummy randomization, excursion drill, identity verification) and short knowledge checks tied to endpoint integrity.
- Delegation and system access drift. Fix by reconciling weekly and requiring PI sign-off for role changes; block activation until matched.
- Consent version and approval threading mismatch. Fix with a visible concordance table and required metadata; prevent “consent in effect” without the matching approval attached.
- UAT without negative tests. Fix by mandating at least one negative test per module with screenshots and expected error messages.
- Unverified pharmacy alarms. Fix with path-task status and logger screenshots filed before IP release.
- Imaging/readiness gaps. Fix with accepted test uploads and pre-booked acquisition blocks where capacity is scarce.
- Courier and hazardous-goods surprises. Fix with a documented test pickup and explicit dry-ice/hazard rules per country before greenlight.
- Decentralized bolt-ons added late. Fix by configuring identity verification, tele-visit scripts, and mobile nurse logistics during SIV, not after FPI.
30–60–90-day implementation plan. Days 1–30: publish checklist and memo templates; finalize thresholds and KRI/QTL definitions; wire click-throughs to eTMF/ISF; define signature meanings; embed country calendars and import/translation lead times. Days 31–60: pilot the checklist at three sites; run retrieval drills; close filing gaps; tune negative-test cases for EDC and IWRS/IRT; perform pharmacy alarm and courier test rehearsals. Days 61–90: scale to all wave-1 sites; institute weekly risk huddles; enforce conditional activation rules; integrate vendor feeds; and review QTLs after the first 10 randomizations per site to adjust early-ramp controls.
Ready-to-use greenlight checklist (paste into your SOP or site initiation pack).
- People: Investigator CV/license current; GCP training; financial disclosures; delegation log complete; knowledge checks passed; system access reconciled to roles (EDC, IWRS/IRT, imaging, eConsent, safety).
- Places: Pharmacy mapping and alarms verified; excursion drill executed; IP receipt and quarantine/release rules filed; imaging/lab test uploads accepted; courier account live with test pickup and hazard documentation.
- Products: Label proofs approved; kit assortment verified; stock reconciled to IWRS/IRT; device/diagnostic configuration recorded with acceptance test pass; quarantine/release rules documented.
- Processes: Ethics/authority approvals on file; localized consent version in effect and threaded to approval; recruitment materials approved; privacy/data-transfer statement filed; eConsent identity verification demonstrated; SAE clock rehearsal completed; emergency unblinding request flow tested.
- Firewalls: Blinded/unblinded roles separated; access approvals filed; code-break and unblinded stock visibility restricted.
- Simulation: First-day table-top/flow test recorded (consent→eligibility→randomization→first dose/procedure), defects logged with owners and due dates.
- Thresholds: Training completion ≥95% (no critical gaps); UAT sign-offs (positive and negative tests) for EDC and IWRS/IRT; courier test pass; identity-verification success ≥98% in two scenarios.
- Conditional Items (if any): Explicit time-boxes and pause rules recorded; owners and due dates assigned.
- Retrieval Drill: Evidence chain produced within five minutes; metadata corrected if drill fails.
- Decision: Go / Conditional Go / No-Go memo signed; each signature includes the meaning of approval; memo filed with links to artifacts.
Bottom line. A greenlight is credible when readiness is demonstrated, not declared. Encode the right criteria, wire them to evidence you can retrieve in minutes, rehearse first-day execution, and run governance that turns signals into dated, documented decisions. With proportionate controls anchored in widely recognized expectations, clear firewalls, and practical early-ramp limits, you will activate faster—and you will be ready to show why you were confident to begin.