Detection and Case Qualification—From Signal to “Day 0” in Minutes

Minimum criteria define a “case.” A valid initial case exists once four elements are present: identifiable patient, identifiable reporter, suspect product, and reportable event. The moment all four exist for the sponsor (or designee), awareness is established and clocks start. Until then, treat reports as signals and pursue missing elements immediately, documenting attempts and next contact times.

Seriousness, relatedness, expectedness—converging quickly and consistently. Intake scripts capture seriousness (outcome-based), provisional relatedness (site/sponsor best-available view), and a first pass at expectedness. Expectedness uses a controlled mapping of MedDRA Preferred Terms (PTs) to the current RSI. Store the RSI version/date with the case; failing to document the reference set in force at onset is a common inspection finding. For marketed background therapies, the governing reference is the local label/package insert recorded in the case metadata.

Temporality and alternative causes—make your reasoning visible. Causality requires a plausible time-to-onset and a transparent search for alternatives (disease progression, intercurrent infection, drug–drug interactions, procedures). For each expedited candidate, write a one-sentence rationale: “Onset X hours/days after dose Y; alternatives explored Z; dechallenge ⬚; conclusion → relatedness at least possible.” A brief, structured justification prevents drift and allows rapid second-reader checks.

Automation with guardrails. Configure the safety database to: (1) trigger “expedited candidate” flags when seriousness + relatedness + unexpectedness are satisfied; (2) pre-populate E2B(R3) fields from structured intake data; (3) lock dictionary and RSI versions on awareness; and (4) block submission status unless a narrative and expectedness reference are attached. Automation accelerates flow but must never replace clinical judgment; type-ahead prompts should guide, not decide.

Duplicates and cross-channel coherence. The same hospitalization can arrive by phone, an EDC SAE form, and a site email. Use deterministic keys (site, subject, onset date) and fuzzy logic (similar terms, same hospital) to merge, link, or supersede entries. Never delete; link and annotate. Duplicate expedited transmissions irritate authorities and create avoidable audit findings.

Decentralized reporting—identity and clock discipline. Tele-visits, eConsent, and home-health introduce identity risk and time-zone complexity. Require two-factor checks for participant-initiated contacts, capture local time plus UTC, and use synchronized server clocks so “awareness” is unambiguous. Courier or home-nursing logs belong in the case packet when they influence onset or seriousness determinations.

Firewalls and unblinding for safety. If causality cannot be reasonably assessed without treatment code, follow a pre-approved minimal-disclosure path that restricts allocation details to a small, unblinded safety unit. Record who was unblinded, why, and when, and keep blinded teams protected. The narrative should state “unblinding performed for safety per SOP” without revealing code to blinded audiences.

Timelines, Transmission, and Follow-Up—Making Deadlines Boring

Build buffers around the regulatory clocks. External timelines for SUSARs are typically measured from “day 0” awareness and include short windows for fatal/life-threatening cases versus other serious unexpected reactions. Design internal service levels to beat external clocks—e.g., validity confirmation within hours; entry and medical review the same day; expectedness mapping and country routing immediately for probable expedited cases; and a submission buffer that anticipates weekends, holidays, and portal outages. Treat after-hours awareness as same-day awareness for internal timing to simplify calculations and narratives.

E2B(R3) quality at the first transmission. A compliant initial transmission includes core ICSR fields (patient/reporter identifiers, suspect product, event term(s), onset course, seriousness criterion, relatedness rationale, expectedness reference/version), a synchronized narrative, and key attachments (e.g., discharge summary, critical labs). Use short “what changed and why” headers in follow-ups to explain updates (new labs, imaging, causality refinement). Never overwrite history—append and explain.

Country distribution lists without guesswork. Maintain controlled distribution logic: study, product, country of occurrence, subject nationality/residency where relevant, and local ethics body rules. Pre-load static data (sponsor contact, product dictionary, trial IDs) in national portals to avoid keystroke delays. Keep language packs and translator SLAs ready for cases requiring translation; vocabulary consistency in medical terms across languages prevents re-work.

Proof matters as much as punctuality. File acknowledgments, receipts, checksum logs, and portal confirmations with the case. Arrange the eTMF so each dashboard date clicks to the narrative, attachments, and proof of submission. During inspection, the first follow-up question after “why was this unexpected?” is “show me that you sent it in time.” Make the answer one click away.

Nullification and corrections—cleanly, with traceability. If a case is later determined not to meet SUSAR criteria (e.g., expectedness flips after RSI update; diagnosis revised), send a correction or nullification per national rules and append a memo that explains the change. Keep a clean chain: original transmission → follow-up(s) → nullification/correction, each with timestamps and owners.

Interfaces and reconciliation. Reconcile safety and EDC frequently: subject ID, onset date, term, seriousness, relatedness, and outcome should match. If the study stores site causality in EDC and sponsor causality in the safety system, label both clearly and declare which governs expedited routing (most conservative plausible view). Document reconciliation cycles and close discrepancies with audit-trailed change notes.

Pregnancy and special populations. Pregnancy exposures and neonatal outcomes often require quicker attention and specific follow-up fields. Ensure the ICSR template prompts for gestational timing, outcome, paternal exposures where relevant, and genetics consults when indicated. For pediatrics or elderly, highlight dose adjustments, comorbidity clusters, and polypharmacy to sharpen expectedness judgments.

Governance, KRIs/QTLs, Pitfalls, and a Ready-to-Use Checklist

Small, named ownership with meaning of approval. Assign an accountable Safety Operations Lead (workflow and submissions), a Safety Physician (causality and medical accuracy), a Data Manager (EDC–safety reconciliation), and Quality (ALCOA++/traceability). Each approval should record its meaning—“medical accuracy verified,” “expectedness reference/version checked,” “country routing confirmed,” “ALCOA++ attributes verified.” Small boards move quickly and are easier to audit.

Dashboards that change behavior. Track: time from awareness to validity confirmation; intake-to-entry and entry-to-medical-review; percentage of cases with narrative/field mismatches at lock; expedited clock burn-down; proportion of interim narratives; translation cycle time; duplicate rate; reconciliation gap rate; and click-through to submission proof. If a number cannot click to an artifact, it is not inspection-ready.

Key Risk Indicators (KRIs) and Quality Tolerance Limits (QTLs). KRIs: spike in “unassessable” causality; repeated RSI mismatches; missing expectedness reference/version; high duplicate rate; portal rejections; or delayed engineering analysis for cases involving device components in combination trials. Convert the most consequential to QTLs—for example, “≥5% expedited cases missing explicit expectedness reference/version in any rolling month,” “≥3 portal rejections in a week,” or “≥10% narrative–field inconsistency at lock.” Crossing a QTL triggers a documented review with containment, correction, and a due-dated action owner.

Training and calibration. Use anonymized, adjudicated examples with a single-fact difference (e.g., onset 2 hours vs. 2 weeks; dechallenge present vs. absent). After any RSI update, run a micro-refresher—expectedness can flip for the same clinical picture when the reference changes. For global teams, add short “how to pronounce the case in plain language” drills so investigator communications are consistent and respectful.

Common pitfalls—and durable fixes.

• Clocks start late because “the report was incomplete.” Fix with the four-criteria rule for case validity and an immutable awareness timestamp.
• Narratives that don’t match coded fields. Fix by generating narratives from structured fields and forcing a pre-lock consistency check.
• RSI/version ambiguity. Fix with a version-locked mapping table and visible RSI metadata at the top of each case.
• Duplicate expedited submissions. Fix with deterministic keys plus fuzzy matching; link rather than delete; teach sites to update the same case.
• Portal rejections at hour eleven. Fix with pre-configured accounts, dry-runs, and a submission buffer; trend rejection reasons and remove root causes.
• Unblinding creep. Fix with a minimal-disclosure path and an unblinded safety unit; log who saw what and why.
• Decentralized identity gaps. Fix with two-factor checks, synchronized clocks, and retention of courier/home-nursing logs where onset or seriousness relies on them.

Ready-to-use SUSAR checklist (paste into your safety plan).

• Minimum case criteria confirmed; immutable awareness (“day 0”) timestamp captured.
• Seriousness and provisional relatedness documented; structured one-sentence causality rationale recorded.
• Expectedness mapped to current RSI and version/date captured; local label used for marketed comparators with version/date.
• MedDRA coding locked to dictionary/version at awareness; narrative synchronized to coded fields; key attachments included.
• Expedited routing logic applied; country distribution list selected; language/translation path ready if needed.
• Submission proof (acknowledgment/receipt/checksum) filed and clickable from the dashboard.
• Follow-up plan time-boxed; “what changed and why” headers used for updates; nullifications/corrections filed when appropriate.
• EDC–safety reconciliation scheduled; discrepancies closed with audit-trailed change notes.
• Unblinding for safety performed via minimal-disclosure path, with access logs retained.
• KRIs/QTLs monitored; red thresholds trigger documented containment and correction with owners and due dates.

Bottom line. SUSAR performance is a design choice, not a lucky outcome. When definitions are stable, expectedness mapping is version-locked, narratives match fields, clocks start on time, and transmission proof is a click away, expedited reporting becomes fast and defensible. Engineer that system once—scripts, mapping tables, buffers, dashboards, and retrieval drills—and you will protect participants, meet deadlines, and sail through inspections across regions and study designs.