What this is
What is a cause code register?
What is a cause code register?
It is the single governed list of cause codes that every investigation in the workspace must pick from, rather than each investigator inventing wording on the spot. Each code sits at a fixed causal level — absent defence, individual action, task condition or organisational factor — and carries a definition, an example and a note on what it must not be used for.
What is the difference between a cause code and a root cause?
A cause code is a reusable label from a fixed list; a root cause is the specific finding written up for one event in a Root Cause Analysis record. The register exists so that separate write-ups can still be counted and compared, because they draw from the same labels.
What is ICAM and the Reason model, and why do they sit together here?
ICAM supplies the four-level structure: absent or failed defences, individual or team actions, task or environmental conditions, and organisational factors. The Reason model is the underlying theory — that accidents happen when latent organisational weaknesses line up with an active failure — which is why the register separates upstream causes from the moment-of-event action.
Scope
When is a cause code register required?
This register holds the taxonomy, not the investigation. Using it to record what happened in a specific event, instead of naming the recurring category it belongs to, pushes narrative detail into a field meant to stay short and reusable.
Use this template when
- You are setting up the workspace for the first time and need a starting taxonomy of cause codes
- An investigation surfaces a genuinely new failure mode that no existing code covers
- Two codes have turned out to mean the same thing and need merging before the next reporting cycle
- The annual review is due and codes need to be confirmed, merged or retired
- Root Cause Analysis needs a governed pick list to draw from rather than free text
Do not use it for
- Root Cause Analysis, which investigates and writes up one incident using codes from this register, not where new codes get invented.
- Just Culture Determination, which decides individual accountability for an action, a separate judgement from naming its causal category.
- Corrective and Preventive Action, which fixes the cause once named, not where the naming happens.
- Standards and Clause Register, which holds external standard clauses, not an internal causal taxonomy.
- Anything outside General, which belongs in the workspace that owns that process
Compliance mapping
Which ICAM requirements does this satisfy?
ICAM and the Reason model don't carry numbered clauses the way a management-system standard does — what they specify is a structure of causal levels and a way of reasoning about them. The requirements below map that structure onto the register's own sections.
| Clause | Requirement | Where it lands |
|---|---|---|
| ICAM — absent or failed defences | Every code must sit at exactly one of the four ICAM causal levels, not float between them depending on who fills the form in. | Identification |
| ICAM — individual or team actions | Codes describing what a person did in the moment must stay separate from codes describing the conditions that set them up to do it. | Identification |
| Reason model — latent conditions vs active failures | Cause Group should sit inside the Cause Level it belongs to, so upstream groups like training never get filed under an active-failure level. | Identification |
| ICAM — cause naming convention | Cause Name is the label every downstream pick list displays, so it must be worded the same way every time, not paraphrased per record. | Identification |
| ICAM — shared definition | Definition must state precisely what qualifies for the code, so a second investigator applies it the same way without asking the author. | Definition |
| ICAM — worked example | Example should anchor the code to a real, recognisable scenario rather than restate the definition in different words. | Definition |
| Reason model — non-overlapping codes | Do Not Use For should name the specific adjacent code this one is most often confused with, not a generic disclaimer. | Definition |
| Register governance — annual review | Added Date, Last Reviewed and Owner must be current, so a code needing merge or retirement doesn't sit untouched for years. | Governance |
What it does not cover
- Cause Name, which duplicates the wording of an existing code under a different label, splitting one failure mode across two entries in every Pareto chart that follows.
- Definition, which restates the code's name in different words instead of stating what qualifies for it.
- Cause Level, which is set to 'Organisational factor' on almost every code regardless of what the code actually describes, because it was the first option in the list.
- Do Not Use For, which is left blank, so investigators keep applying the code to the exact case it was meant to exclude.
- Owner, which names a department rather than a person, so no one is ever the one who missed the annual review.
Global
Cause Code Register requirements by country
Neither ICAM nor the Reason model is a legal instrument, so 'jurisdiction' here means where the expectation to use a governed cause taxonomy actually comes from.
ICAM, as promoted by state work health and safety regulators and the resources sector
ICAM is most heavily codified across Australian mining, rail and energy regulators, who expect a documented causal taxonomy behind serious-incident investigations.
An Australian site with only free-text investigation narratives, no governed code list, is a step behind its own regulator's stated good practice.
HSE investigation guidance drawing on the Reason organisational accident model
UK guidance leans on Reason's split between latent organisational conditions and active failures when judging whether an investigation went deep enough.
A UK investigation that stops at the worker's action, with no coded route back to an organisational factor, reads as incomplete against that guidance.
ISO 45001 root cause and non-conformance expectations
ISO 45001 doesn't mandate ICAM specifically, but its requirement to determine causes of nonconformity is commonly satisfied with a coded taxonomy like this one.
An auditor will accept a governed register as evidence of a systematic method; a pile of one-off explanations usually draws a finding.
How to complete it
How to complete a cause code register, step by step
Filling in the fields is mechanical. The choices deciding whether the register produces a usable Pareto chart a year from now are these four.
Too broad and every finding lands on 'communication', telling you nothing you didn't suspect. Too narrow and the list grows past the point anyone can browse it before typing a new one. Aim for a handful of codes covering most of a year's investigations.
Cause Level gives the ICAM four-way split; Cause Group gives the practical bucket (design, training, supervision) that maintenance and HR act on. Pick the primary reporting axis before codes are added at scale.
If any investigator can create a code mid-investigation, the register stops being governed within a quarter. Naming the approving owner before the first request comes in is what keeps this a taxonomy, not a commentary.
Merging preserves the historical count under the surviving code; retiring without a merge just hides the events that used it. Recording the mapping in Do Not Use For keeps historical charts comparable.
What auditors find
Most common cause code register findings
These are the patterns that show up most often once a cause code register has run for a year without a firm owner.
| Finding | Clause | What fixes it |
|---|---|---|
| Two active codes describe the same failure mode under different names. | ICAM — cause naming convention | Merge into one code, redirect the Root Cause Analysis pick list, and record the retired name in Do Not Use For. |
| Cause Level and Cause Group disagree — a code marked 'Organisational factor' sits in the 'Training' group but reads like an individual action. | Reason model — latent conditions vs active failures | Re-read the Definition and reassign whichever field doesn't match what the code describes. |
| Definition is a one-line restatement of Cause Name with no added meaning. | ICAM — shared definition | Rewrite so a second investigator could apply the code without asking the author what it meant. |
| Example is blank on most codes in the register. | ICAM — worked example | Pull one real instance from a closed Root Cause Analysis record and paste in a short, specific version. |
| Owner is a shared team name rather than a named person. | Register governance — annual review | Assign an individual; a team can't be asked why a stale code wasn't reviewed. |
| Last Reviewed is over a year old on codes still marked Complete or Closed. | Register governance — annual review | Run the overdue codes through the annual review and confirm, merge or retire each one. |
Case in point
Case in point: the register that grew a new code every month
A site ran root cause analysis diligently for a year with no one gatekeeping the code list. Every investigator who couldn't find quite the right wording typed a new Cause Name into the linked pick list. By the annual review there were 61 codes for what a fresh read-through showed was really a dozen distinct failure modes — half of them near-duplicates like 'Inadequate supervision' and 'Supervision gap', each with no Definition explaining how they differed.
The fix was an owner, not a new field. The safety lead became the sole person who could add a code, near-duplicates were merged with the old names kept in Do Not Use For, and surviving codes got a real Definition and Example. The next Pareto chart showed three failure modes driving most of the year's incidents — visible for the first time because the codes describing them were finally counted together.
The template
The template, field by field
The form exactly as it installs. Every field, option, score and conditional rule is editable, and the links to other templates come with it.
3 sections
- Reference
- FDN-010
- Archetype
- Register
- Record ID
- CAUSE-2026-000
- Scoring
- None
- Direction
- n/a
- Singleton
- Yes
- Basis
- ICAM, Reason model
- Links
- Used by Root Cause Analysis
- Tags
- Master data, Investigation
- Sections
- 3
- Fields
- 11
- Follow up fields
- 0
- Repeating sections
- 0
- Links out
- 0
Identification
5 fieldsCause Code*
Format CAUSE-000.
The record's own ID. Other templates point at this value.
Status*
- Planned2 pts
- In progress2 pts
- Complete3 pts
- Deferred0 pts
- Open0 pts
- Closed3 pts
- Overdue0 pts
Cause Level*
Failed defence, individual action, task condition or organisational factor.
Cause Group*
Groups codes so a Pareto chart is readable.
Cause Name*
This is what Pick Lists display.
Definition
3 fieldsDefinition*
What this code means, so two investigators use it the same way.
Example
A short real example makes the code far easier to apply consistently.
Do Not Use For
The common misuse, named explicitly.
Governance
3 fieldsAdded Date
Last Reviewed
Owner
FDN-010 · record IDs look like CAUSE-2026-000 · Used by Root Cause Analysis
Open in KnowellaRun it with agents
From a document you fill in to a programme that runs itself
The register itself is short. What actually determines whether it stays governed is whether someone stops a duplicate code before it's created, and whether the annual review happens on time.
Checks a proposed Cause Name against the existing register before it's added, flagging likely duplicates before a new code gets created for a failure mode that already has one.
Rolls Root Cause Analysis records up by Cause Group into a running Pareto view, so drift toward one over-used code shows up before the annual review.
Tracks Last Reviewed against the annual cycle and surfaces stale codes, so the review is a short confirmed list rather than a blank-page rebuild.
Holds the register against every linked Root Cause Analysis record, routes merge or retire decisions to the named owner, and keeps the evidence trail together.
This template lives in General — control tower. The orchestration layer. Registries and engines every other workspace reads from.
Meet General→Glossary
Cause Code Register definitions and key terms
- ICAM
- Incident Cause Analysis Method — an investigation framework that sorts causes into four levels: absent or failed defences, individual or team actions, task or environmental conditions, and organisational factors.
- Reason model
- James Reason's organisational accident model, often called the Swiss cheese model, which explains accidents as the alignment of latent organisational weaknesses with an active failure at the point of harm.
- Latent condition
- A weakness built into a system well before an event — a design flaw, a gap in training, an unrealistic schedule — that sits dormant until it lines up with an active failure.
- Active failure
- The action or omission at the moment of the event, usually the most visible cause and the easiest one to blame without looking further upstream.
- Pareto chart
- A ranked bar chart showing which cause codes account for most investigations, so effort goes toward the few categories driving most of the harm.
FAQ
Frequently asked questions about cause code register
Is the cause code register the same thing as a root cause analysis?+
No. This register holds the reusable taxonomy; Root Cause Analysis is the record for one investigation, which selects codes from here rather than inventing its own wording.
How many cause codes should a workspace start with?+
Most teams start with fifteen to thirty codes spread across the four ICAM levels, then add sparingly. A register past sixty or seventy usually has duplicates, not genuinely new failure modes.
Who should be allowed to add a new code?+
One named owner, usually the safety or quality lead, not every investigator. Open authorship is the most common cause of the near-duplicates that make reporting unreliable.
What goes in Do Not Use For if nothing obvious is being confused yet?+
Leave it until confusion happens, then name the specific adjacent code investigators keep reaching for instead. A generic disclaimer written in advance rarely names the real overlap.
Does Cause Level or Cause Group decide how a code shows up in reporting?+
Both are recorded, but pick one as the primary axis before codes are added at scale — Cause Level for the ICAM split, Cause Group for an actionable bucket like training or supervision.
Can old codes be deleted once they're no longer used?+
Retire rather than delete, and note in Do Not Use For whether it was merged into a survivor. Deleting a code that historical records reference breaks last year's Pareto chart.
Keep going
Related templates and programmes
Industries this is written for
Programmes this belongs to
Used together in Master Data and Foundations
Site and Location Register
Holds every site, building, area and zone your organisation operates
Asset Register
Holds every piece of equipment, machine, vehicle and tool you track
Worker Profile
Holds a record for each worker, including role, department, site and start date
Job and Task Register
Lists the jobs and tasks people perform, so risk assessments and ergonomic assessments can be tied to real work rather than job titles
Vendor and Contractor Register
Holds every supplier, contractor and service provider you work with, including their status and approval level
Chemical and Substance Register
Lists every chemical and hazardous substance held on site, with quantity, location and hazard class
More in Registries
Site and Location Register
Holds every site, building, area and zone your organisation operates
Asset Register
Holds every piece of equipment, machine, vehicle and tool you track
Worker Profile
Holds a record for each worker, including role, department, site and start date
Job and Task Register
Lists the jobs and tasks people perform, so risk assessments and ergonomic assessments can be tied to real work rather than job titles
Vendor and Contractor Register
Holds every supplier, contractor and service provider you work with, including their status and approval level
Chemical and Substance Register
Lists every chemical and hazardous substance held on site, with quantity, location and hazard class
Written and reviewed by
Siddarth Singh
Founder & Chief Executive Officer, Knowella
Certified Safety Professional and industrial and systems engineer with more than a decade inside food supply chain, freight and manufacturing operations. This page was written against the current text of the standards it cites, not against secondary summaries of them.
- Certified Safety Professional (CSP), Board of Certified Safety Professionals
- MBA, University of Chicago Booth School of Business
- MS and BS, The Ohio State University, Industrial and Systems Engineering
- Six Sigma Black Belt
Sources and last review. Reviewed 16 August 2026 against:
- ICAM — Incident Cause Analysis Method
- Reason organisational accident model (Swiss cheese model)
- ISO 45001 — cause determination for nonconformity
This page is general guidance, not legal advice. Confirm requirements with your jurisdiction’s regulator.