Adverse Event Records

This topic is part of the SG Systems Global Guides library for regulated manufacturing teams evaluating eBMR, MES, and QMS controls.

Updated December 2025 • adverse event records, serious adverse event (SAE) workflow, record retention, privacy controls, MedWatch, follow-up reporting, complaint vs AE triage, audit trails • Dietary Supplements (USA)

Adverse event records are the controlled evidence set you maintain when a consumer, healthcare professional, retailer, or distributor reports a health-related experience associated with your dietary supplement. These records are not the same thing as marketing feedback, customer service tickets, or normal product complaints. They are a regulated, high-stakes body of information that must be captured consistently, protected for privacy, linked to lot and distribution traceability, and reviewable under inspection conditions without turning into chaos.

Teams usually discover they have weak adverse event records at the worst possible time: after a serious case, when legal or regulatory pressure escalates, or when a customer asks “show me your adverse event program.” The solution is not just to “store emails.” The solution is to build a system where adverse event intake, follow-up, medical review, seriousness determination, reportability decisions, and record retention are governed like any other critical quality workflow. For supplement operations context, see Dietary Supplements Manufacturing.

“The fastest way to lose control is to treat adverse events like customer service. The fastest way to regain control is to treat them like evidence.”

1) What buyers mean by “adverse event records”

Buyers aren’t asking for “a folder where we put stuff.” They’re asking for a complete, defensible case file that can answer hard questions quickly: What happened? To whom? When? What product and lot? What medical outcome? What follow-up occurred? Was it serious? Was it reportable? What did the company do next? Which other lots might be affected? Who reviewed and approved decisions? Can you prove the record hasn’t been altered?

In practical terms, adverse event records are a specialized branch of complaint trending and post-market surveillance, with higher privacy requirements and tighter time management. They also overlap with quality events (deviations, OOS, nonconformance) when the underlying issue is a product defect. A mature system keeps these threads distinct but linked: you can’t manage what you can’t connect, and you can’t defend what you can’t separate.

2) Complaint vs adverse event vs product quality issue

This separation is the first control point. Many companies lose time and credibility because they treat everything as a “complaint” until it becomes serious. The correct model is triage with clear routing rules:

In the real world, these can overlap. A complaint can contain adverse event information (“I got sick”). An adverse event can reveal a product quality issue (contamination). Your system must support “linked cases” without mixing the privacy profile of AE data into general complaint handling. This is where role-based access and structured forms matter.

3) What “serious” means operationally and why it changes everything

“Serious” is not a feeling. It is a classification with operational consequences. You need a clear seriousness decision tree with documented rationale. In practice, seriousness determines:

• How fast you must complete review and submission workflows
• Whether senior medical/QA review is mandatory
• Whether immediate distribution impact assessment is required
• Whether you need to implement containment actions (holds, retailer notifications, recall readiness)

A mature program defines seriousness criteria in the system as structured fields and ensures the decision is reviewable. It also separates “seriousness” from “causality.” A case can be serious even if causality is uncertain. GDP discipline matters here: your record must show what you knew at the time, what you did, and why.

4) Intake channels: phone, email, web forms, retailers, and medical professionals

Your intake channel is not just “how people contact you.” It determines whether you capture the right minimum information early and whether the clock starts cleanly. Common intake channels:

• Label phone number/email (direct consumer contact)
• Website contact form (often routed to customer support first)
• Retailer portals (Amazon, big-box retailers, distributors)
• Healthcare professional contact (sometimes via email or fax)
• Social media (high noise, but can contain real AE info)

Best practice is to route any health-impact statement into a controlled AE intake form immediately—without losing the original communication. That means the AE record stores: the original message (unaltered), the structured intake fields, and a link to any complaint or quality incident record created in parallel.

Operational tip: use a “two-layer form.” Layer 1 captures just enough to create a case and start follow-up (reporter contact, product, event description). Layer 2 is the deep medical and traceability capture. If you force every caller into a 25-minute interview, they will hang up. If you capture too little, you’ll chase missing data later. Balance is design.

5) Minimum data elements: what you must capture on day one

AE records live or die on “minimum viable completeness.” You cannot build a serious-case submission or a defensible internal review if you don’t capture core identity and outcome elements early. A practical minimum set includes:

Notice what isn’t required on day one: a perfect medical history or a definitive causality statement. Your goal is to capture enough to (1) classify seriousness, (2) preserve product identity, and (3) enable follow-up. Everything else is staged through the workflow.

6) Product identification: label, SKU, lot, expiry, and “which version”

Most AE records fail at product identity. Consumers don’t speak in SKUs. They say “the green bottle.” Retailers store data differently. Labels get revised. Claims change. You need a product identity strategy that tolerates real-world ambiguity.

Practical controls:

• Capture label images early (front panel + supplement facts + lot/expiry area). Store as immutable attachments with audit trails.
• Normalize product mapping to internal SKU/UPC and label revision (tie to revision control).
• Capture lot/expiry as structured fields and also as photo evidence. A typed lot without photo is fragile.
• Handle “unknown lot” cases explicitly with escalation rules (ask for photo, purchase receipt, retailer order number).

This is also where your lot numbering discipline pays off. If your lots are ambiguous, you can’t scope impact quickly. Link to Lot Numbering Strategy if/when you publish it.

7) Lot linkage and traceability scope: from complaint to impacted shipments

AEs aren’t just “medical.” They are operational. When a credible case arrives, you need to answer: is this isolated or systemic? That requires linking to traceability.

At minimum, your system should support:

• Lot genealogy (which components, which suppliers, which MMR/BPR version) via end-to-end genealogy.
• Distribution scope (which customers/regions received the lot) for fast containment decisions.
• Related quality events (deviations, OOS/OOT signals) tied to the same lots.
• Reserve samples linkage when you need confirmatory testing (tie to a controlled sample record and custody).

The fastest way to lose days is to do traceability manually after a serious case arrives. If you can click from AE case → finished lot → component lots → supplier lots → shipments, you can do impact assessment while your medical reviewer is doing follow-up.

8) Medical review: narratives, follow-up, causality, and documentation discipline

Most supplement companies are not running a pharmacovigilance department, and they shouldn’t pretend they are. But you still need a disciplined medical review workflow that is defensible. That means:

• Structured narrative capture. Keep the consumer’s words as an original record, then add your structured interpretation separately.
• Follow-up questions that are consistent: timing, dose, product identity, medical intervention, outcome, other products. Use templates so two reviewers don’t ask completely different things.
• Document “unknowns.” If the reporter refuses data or disappears, record what you attempted and what you could not obtain. Silence is not the same as “no issue.”
• Separate seriousness from causality. Seriousness is outcome-based; causality can be “unknown” and still defensible if your rationale is clear.

GDP matters here more than anywhere else because AE narratives are messy. Your goal is not to force a clean story. Your goal is to preserve what was reported, preserve what you did next, and preserve how decisions were made. If your AE records show “we concluded not related” with no rationale, you’ve created risk.

9) Privacy and sensitive data: access control, redaction, and data minimization

Adverse events can contain sensitive personal data: medical details, diagnoses, medications, contact information, and sometimes documents. Regardless of whether your company is a “covered entity,” the operational expectation is the same: treat this data as sensitive and minimize exposure.

Practical privacy controls that work:

• Role-based access. Only trained roles can view full AE case details (RBAC). Customer support should see only what they need to route the case.
• Field-level sensitivity. Separate “identity/contact” from “medical narrative” from “internal assessment.” Not everyone needs all layers.
• Redaction workflow. If you must share records with customers or partners, provide a controlled export path that can redact personal identifiers while keeping evidence intact.
• Data minimization. Don’t collect unnecessary medical data. Collect what you need to assess seriousness, product linkage, and follow-up.
• Auditability. Any access to sensitive data should be logged (who viewed, when) where feasible—especially in high-risk cases.

10) Timelines: triage clocks, follow-up clocks, and submission clocks

AEs create clocks. If you don’t time-manage them, you will eventually miss a required action or you’ll discover a gap when you can’t fix it. A robust system tracks at least three timing layers:

• Intake clock: when the report was received (the “received date” is a control field, not a guess).
• Triage clock: when seriousness determination must be completed (internal SLA).
• Submission clock: if reportable, when submission must occur and when follow-up must be sent.

Even for non-serious cases, internal clocks matter: follow-up attempts, closure criteria, and trending review cadence. For serious cases, the system should behave like a dispatch tool: required tasks, responsible owners, and escalation if deadlines approach. “We thought someone else was doing it” is not defensible.

11) Follow-up records: new medical info, duplicate cases, and consolidation

Adverse event records are not one-and-done. They often evolve: the consumer goes to urgent care, symptoms resolve or worsen, lab results appear, a hospital report is provided, a second caller reports the same event, or a retailer forwards additional information later.

Your system must support:

• Follow-up entries that preserve the original record and add new information as versioned, time-stamped additions.
• Duplicate detection (same reporter + product + timeframe) so you don’t double-count signals or miss consolidation requirements.
• “New medical information” linkage back to the original serious case so follow-up submissions can be time-managed.
• Case consolidation where the system can show a master case with linked duplicates, without losing the original messages.

This is where weak programs collapse: they store the “final story” but lose the chronology. Chronology is evidence. If you can’t show the timeline of what was known when, your record is not defensible.

12) Recordkeeping and retention: what to store, how long, how to retrieve

AE records should be treated as an evidence pack. Even if your team is small, the record must be complete enough that someone unfamiliar with the case can reconstruct the decision path quickly. A practical record pack includes:

• Original intake (email, call notes, web submission, retailer message)
• Structured case form (product, lot, dates, seriousness classification)
• Attachments (label photos, receipts, medical documents if provided)
• Follow-up attempts and outcomes (including “no response”)
• Internal medical/QA assessment notes (as controlled entries)
• Submission copy and confirmation (if reportable)
• Any linked quality events (deviation/OOS/label issue) and resulting actions
• Trending classification and closure rationale

Retention must be managed like any other critical record. Your record retention and archival posture must ensure the case is retrievable years later. “We had it on someone’s laptop” is not acceptable. Records should be stored in a controlled system with access control, backups, and auditability.

13) Inspection readiness: what FDA or customers will ask to see

Inspection readiness is not “we can find the PDF.” It’s “we can answer questions fast and consistently.” Expect scrutiny on:

• Process clarity: documented procedures for AE intake, seriousness determination, reporting, follow-up, and retention.
• Completeness: cases show minimum data elements, follow-up efforts, and clear closure rationale.
• Timeliness: serious cases show a clear received date, triage date, and submission date with no ambiguity.
• Data integrity: audit trails show changes, reasons, and approvals; no silent edits.
• Linkage: ability to connect cases to lots and distribution and to show impact assessment when required.
• Trending: ability to show case counts by product, lot, symptom category, seriousness, and time window.

If you can generate a clean “AE program packet” on demand—SOPs, case logs, trending, and selected case evidence—you’ll be in control. If you need a week to reconstruct it, you’re not ready.

14) Trending: signal detection, thresholds, and escalation into QMS events

Trending is where AE records become proactive. A single AE rarely proves anything. Patterns do. Your trending model should include:

• By product/SKU and label revision (claims and formulation changes matter)
• By lot (cluster detection is critical)
• By symptom category (standard categories + free-text narrative preserved)
• By seriousness (serious vs non-serious signals behave differently)
• By channel (retailer vs direct; geography; seasonality)

Trending should drive thresholds. Example thresholds (you must tailor to your risk appetite):

• Two serious cases associated with the same finished lot → immediate quality impact assessment + hold distribution if possible
• Unusual symptom cluster for a product over 30 days → open investigation even if cases are non-serious
• AE signals correlated with supplier change or formulation revision → escalate to change control review
• Pattern of “unknown lot” AEs → improve labeling clarity, intake questions, or retailer data capture

Once thresholds are defined, the system should enforce escalation into QMS workflows: deviation investigation, CAPA, or supplier corrective action where appropriate. Without escalation rules, trending becomes reporting theater.

15) Actions: when AE records trigger holds, recall drills, or CAPA

AE records are not just records—they drive decisions. The decision set typically includes:

• No action (document rationale; continue monitoring)
• Enhanced monitoring (tighten sampling, increase follow-up, trend review cadence)
• Quality investigation (link to lots, review batch records, review lab results, inspect retains)
• Supplier escalation (change notification check, COA verification, SCAR if evidence supports)
• Distribution containment (hold remaining inventory, notify distribution partners as needed)
• Recall readiness drills (simulate scope and response) — see Recall Drill and Recall Readiness
• CAPA for systemic issues (repeat patterns, weak controls, recurring documentation gaps)

The strongest organizations treat AE handling as a closed-loop system. Intake and records are the front door. Trending and action are the payoff. If your AE records don’t feed learning, you will relive the same crisis repeatedly.

16) Copy/paste demo script and selection scorecard

Use this demo script to force vendors (or internal teams) to prove the AE workflow is executable and defensible.

17) Selection pitfalls (how AE programs fail under pressure)

• Inbox-as-system. AE data scattered across emails and spreadsheets becomes unrecoverable under stress.
• Mixed privacy zones. Customer support sees medical data they don’t need; exports leak sensitive information.
• No lot evidence. Cases lack lot/label photo early, so impact assessment becomes guesswork.
• Chronology collapse. “Final story” replaces the timeline; you can’t prove what you knew when.
• Silent edits. Without audit trails, credibility collapses immediately.
• Weak seriousness logic. Seriousness determined inconsistently; cases escalated late.
• No trending thresholds. Patterns exist but do not trigger action; you stay reactive.
• Records not retrievable. You can’t produce a complete evidence pack during inspection or customer audit.

18) How this maps to V5 by SG Systems Global

V5 supports adverse event record governance by combining controlled QMS workflows with traceability and data integrity controls—so AE cases link to lots, actions, and evidence without exposing sensitive data broadly.

• Governance and workflow: V5 QMS supports case intake, triage, investigation linkage, approvals, and CAPA triggers with audit-ready records.
• Traceability and scope: V5 WMS supports lot genealogy and distribution linkage for fast impact assessment and containment.
• Execution evidence: V5 MES provides batch evidence context when AE trends suggest a quality mechanism.
• Integration and portals: V5 Connect API supports structured intake sources and integration to external reporting or customer service systems.
• Industry fit: Dietary Supplements Manufacturing.
• Platform view: V5 solution overview.

19) Extended FAQ

Q1. What should be in an adverse event record?
At minimum: who reported it, what happened, when it happened, what product/lot was involved, what outcome occurred, what follow-up you performed, and how decisions were made.

Q2. Do we need lot numbers for every case?
You should try. A case without lot evidence is hard to trend and hard to scope. Capture label photos and receipts early to reduce “unknown lot” dead ends.

Q3. How do we keep AE data private?
Use role-based access, separate sensitive fields from general case fields, log access where possible, and use redacted exports for external sharing.

Q4. Should non-serious AEs be retained and trended?
Yes. Non-serious cases are often the earliest signals. Trending them by product, lot, and symptom category is how you detect drift early.

Q5. What’s the biggest AE recordkeeping failure?
Losing chronology—overwriting narratives or consolidating cases without preserving the original record and follow-up timeline.

Related Reading
• Supplements Industry: Dietary Supplements Manufacturing
• Core Guides: Complaint Management Software | Complaints & Recalls Hub | Recall Readiness Software | Audit Trail Software | Electronic Signatures (Part 11) | Good Documentation Practices
• Quality Workflows: Deviation Management | CAPA for Dietary Supplements | OOS Investigation | Review by Exception
• Glossary: Complaint Trending | Chain of Custody | MedWatch Form | Record Retention | Role-Based Access | Audit Trail (GxP)
• V5 Products: V5 Solution Overview | V5 QMS | V5 WMS | V5 MES | V5 Connect API