Incident Investigation and Learning Program

This program is the department’s way of turning what happens on the fireground, in the station, on the road and in training into knowledge that changes how the work is done. It is working when every injury, exposure, near miss, good catch and equipment failure is reported through a route people trust; when each event is sorted within a day into one of three review tiers by a group that includes the workers’ side; when reviews are led by trained people who ask how the situation made sense to those in it before asking who did what; when every recommendation has an owner, a date and a strength rating, and the chief records why each one was accepted, rejected or deferred; when lessons reach every station in the quarter; and when the number of near-miss reports goes up in the first year because members have learned that reporting is safe and that something happens. The program teaches investigation and runs After-Action Reviews and Post-Incident Analyses itself. For the most serious events it arranges an Independent Learning Review. It never lets learning information be used for discipline.

Every department, every stage. This is one of the four programs a department stands up first, because it is the loop that makes every other program improve and because the statutory duties it carries (reporting, scene preservation, committee participation in investigation) apply from the first day. Profile Worksheet question: none; the module is always on. The department sets the scale: a volunteer department with a single station runs the same three tiers with a triage group of three people who confer by phone.

• Event: an incident, near miss, good catch, exercise, equipment failure, work refusal, or a response that went well and should be studied (The Human Factor, 2026a).
• Near miss: an event that did not cause harm but could have under slightly different conditions.
• Good catch: a hazard, defect or error noticed and corrected before it became an event.
• Exposure: contact with products of combustion, a hazardous substance, a biological agent, excessive noise, heat or cold, or a traumatic event, whether or not harm is apparent.
• Event triage: the sorting of a reported event into a review tier within 24 hours, by a Triage Group, against written triggers (The Human Factor, 2026a).
• Tier 1 After-Action Review (AAR): a short, structured crew review within 72 hours using four questions.
• Tier 2 Post-Incident Analysis (PIA): a facilitated review by trained people who were not at the event, with terms of reference, a timeline, learning interviews, systems analysis and rated recommendations, within 30 days (60 at most).
• Tier 3 Independent Learning Review: a review led by an independent reviewer from outside the department under joint oversight that includes labour, with a public summary; it runs alongside statutory processes and is never called an investigation.
• Learning firewall: the written commitment that information gathered for learning is not used for discipline, performance management or any adverse action (The Human Factor, 2026a; The Human Factor, 2026b).
• Accountability process: the separate route, run by different people under the conduct policy and collective agreement, for conduct concerns: intentional harm, impairment, reckless disregard, criminal acts.
• Local rationality: the principle that people’s actions made sense to them at the time given what they knew, what they were trying to do and the pressures on them (Woods & Cook, 1999; Dekker, 2014).
• Contributing factor: a condition in the work system that shaped what happened, classified by work-system element (Holden et al., 2013)DOI.
• Hierarchy of controls: the ordering of controls from elimination and substitution through engineering and administrative controls to PPE, used to rate recommendation strength (NIOSH, n.d.-d).
• Corrective Action Register: the single record in which every deficiency and recommendation is tracked to verified closure (Corrective Action Register).

R = responsible, A = accountable (one per row), C = consulted, I = informed.

The senior officer responsible for health and safety is the program manager for this program; where the department has a Health and Safety Program Improvement Team, that team owns the program’s content. Full role text: 01 Roles and Responsibilities.

Requirement. Every injury, exposure, near miss, good catch, equipment failure and work refusal is reported before the end of the shift or, for a volunteer who has left the station, within 24 hours. Anyone may report anything, including a response that went well and should be studied. This element answers the accreditation criteria on hazard reporting and near-miss reporting (CPSE, 2019, criteria 11A.2 and 11A.6) and the standard’s requirement to investigate near misses as well as injuries (NFPA 1550, 2024, ch. 6).

What counts.

Minimum method.

• A 60-second path for near misses and good catches: a QR code on every apparatus and station wall leading to a six-field form, a paper copy of the same form in every apparatus, and a phone number answered by the on-duty district chief. A member may report orally to any officer, who records it. Anonymous reports are accepted. The design follows the volunteer research finding that people report when it is simple and they know what happens as a result (NZFS, 2016b).
• The Form - Injury and Incident Report for injuries, exposures and equipment failures, completed by the member (or by the officer if the member cannot) with a supervisor’s section.
• Exposure to products of combustion is reported per fire by the company officer for the whole crew on the exposure log in Decontamination and Contamination Control Program; the individual exposure report is used for anything outside that routine.
• The company officer sends every report to the health and safety officer and the district chief the same shift. The HSO acknowledges every report to the reporter within five working days and tells the reporter what happened to it within 30 days.
• The definition table above, with examples, is posted in every station and taught in recruit and officer training.

Evidence. Reports in {{Records system}}; acknowledgement log; near-miss and good-catch count per 100 calls and per working fire, by station, reported quarterly.

Stage. S1 for injuries, exposures, refusals and statutory events; S1 for the near-miss path; S2 for the acknowledgement and feedback clocks measured.

Requirement. The department reports to the regulator and the workers’ compensation board within the statutory times, preserves the scene of a serious event until released, and notifies the JOHSC, as the applicable OHS statute requires (J08 Incident Reporting and Investigation).

Minimum method. The senior officer responsible for health and safety holds the reporting procedure, written from the department’s own jurisdiction section of J08 Incident Reporting and Investigation: who calls whom, from what number, within what time, and what is written down. Any officer who becomes aware of a death, a likely fatal injury, or an event on the statutory list calls the senior officer immediately, day or night; the senior officer makes the regulator call and records the time, the officer spoken to and what was said. The incident commander preserves the scene except to the extent needed to rescue, treat, or prevent further harm, and records what was moved. The JOHSC co-chairs are notified the same day. A copy of the WCB notice, where the law allows it to serve as the regulator notice, is filed with the call log. The incident safety officer begins the on-scene part of the investigation within the incident area (NFPA 1550, 2024, s. 5.6).

Evidence. Call log; copies of notices; JOHSC notification record. The clocks and the list of reportable events differ by jurisdiction (J08 Incident Reporting and Investigation).

Stage. S1.

Requirement. Every reported event is sorted into a tier within 24 hours or by the next business day, by a Triage Group, against written triggers, with the decision recorded (The Human Factor, 2026a).

The Triage Group. For worker-safety events: the senior officer on duty, the health and safety officer, and the JOHSC worker co-chair or designate. Each seat has a named backup so that triage happens on weekends and at volunteer stations. The group confers by phone or message; it does not need to meet.

Rules.

1. The highest trigger met sets the tier.
2. Any member may ask for a higher tier, and the request is recorded with the decision.
3. De-escalation below the tier the triggers indicate needs the whole group’s agreement, with reasons recorded.
4. If the group disagrees, the higher tier stands.
5. Clear Tier 1 events (a routine working fire, a drill, a good catch with an obvious fix) are confirmed on a shared daily log rather than individually conferenced.
6. For a Tier 2 event the sponsor and the lead facilitator are named within 72 hours. For a Tier 3 event the fire chief is informed immediately and the independent reviewer is engaged within five working days.
7. Triage checks three things before it closes: statutory reporting done; the people involved offered support; the scene and records preserved where a higher tier or a statutory process may need them.

Evidence. Form - Event Triage Record for every event above a clear Tier 1; the daily log for clear Tier 1 events.

Stage. S1.

The tiers are The Human Factor’s After the Call model (The Human Factor, 2026a), restated here so the department can run them from this note. Where the department wants the full method, including the worked examples, sector equivalents and the decision flow, see Manual 02 - The Three Tiers and Event Triage.

Tier 1 triggers. Significant calls; every working fire; multi-unit or multi-agency responses; exercises and live-fire training; good catches; any event where anyone asks for one.

Tier 2 triggers. A near miss with the potential for death or critical injury, such as a mayday, a collapse close call, a lost member, a fall arrested, or an out-of-air event; a lost-time injury; any event on the statutory reporting list; an emergency vehicle collision with injury or significant damage; a significant operational failure (accountability lost, water supply failed, command not established, RIT not in place at a working fire); patient harm during medical response; a workplace violence injury; a work refusal or stop-work call that was not resolved at the scene; a repeated Tier 1 issue.

Tier 3 triggers. A death or critical injury; an event in which senior leaders’ decisions are within scope; low labour-management trust, or either side requests independence; multiple agencies with shared responsibility; major external scrutiny expected (coroner, regulator prosecution, media, council inquiry); recurrence of an event after earlier recommendations; no independent Tier 2 capacity in the department.

Rules that apply to every tier (The Human Factor, 2026a; Dekker, 2014; Hollnagel, 2014).

• Care for people first. Support is offered before any interview and does not depend on taking part.
• A review looks for how the work made sense to the people doing it, and for the conditions that shaped their choices, before it looks at what they should have done.
• A review records what went well and why, as well as what went wrong.
• Every tier produces recommendations that go to the Corrective Action Register; none produces a finding about an individual.
• Participants are briefed on the findings before anyone else is.

Stage. Tier 1 and the Tier 2 method S1; two trained facilitators S1; Tier 3 arrangements in place S2.

Requirement. Information gathered for learning (near-miss and good-catch reports, learning-team write-ups, AAR sheets, learning interview notes, PIA and learning review reports) is not used as the basis for discipline, performance management, or any adverse employment action, and is not placed in a personnel file. The firewall exists because a culture of blame suppresses reporting (Reason, 1997; Dekker, 2016), and because the response to the last report decides whether there is a next one (Conklin, 2019).

The exception, and how it is handled. A conduct concern (intentional harm, impairment on duty, reckless disregard for a known rule that put others at risk, or a criminal act) goes to a separate accountability process run by different people under the department’s conduct policy and the collective agreement; the distinction between error, at-risk behaviour and reckless behaviour follows the just-culture literature (Marx, 2001; Dekker, 2016). If a review team encounters such a concern it stops, informs the sponsor, and the sponsor refers the concern through the separate route using only the fact that a concern exists, not the learning interview content. The learning review continues on the systems questions. The fire chief decides any such referral and records the decision.

For clarity. The firewall does not prevent the department from acting on a hazard; from meeting a statutory duty; from complying with a regulator’s or a court’s lawful demand for records; from disclosing de-identified lessons; or from managing performance through the ordinary supervisory route using information that was independently available to the supervisor. It protects the learning information, not the conduct.

How members know. Every participant in a Tier 2 or Tier 3 review receives the firewall statement in writing before the interview. The near-miss form carries it. The policy clause is counsel-reviewed and agreed with the bargaining agents and the volunteer association (Policy - Incident, Injury, Exposure and Near-Miss Reporting and Investigation clause 8).

Evidence. The signed policy; the written notice template in the toolkit; the JOHSC’s annual check that no discipline has rested on learning information.

Stage. S1.

The method follows the national incident investigation standard (CSA Z1005:21, cl. 6) and the human-factors literature, in the department’s own words. It is used in full for Tier 2 and adapted by the independent reviewer for Tier 3; Tier 1 uses the four questions only. It is the department’s process to investigate and document accidents and injuries to determine root cause (CPSE, 2019, criterion 11A.7), with the difference that this package treats “root cause” as the set of contributing conditions in the work system, not a single failure.

1. Trained people. Every Tier 2 is facilitated by someone who holds the department’s investigator training (section 6) and was not part of the event. The team includes a JOHSC worker member and, where the event was operational, an officer with current command qualification who was not at the scene. The sponsor signs the terms of reference.
2. Terms of reference. Scope, team, what is in and out, timeline, access to records and people, the firewall statement, the report audience, and the statutory processes running alongside.
3. Preservation and collection. Dispatch and radio records, the accountability record, the incident action plan, pre-plans, apparatus telematics, body-worn or helmet camera where used, weather, photographs, equipment set aside and tagged, training and inspection records, and the exposure log.
4. Timeline. A single timeline of what happened, built from records first and interviews second, with the time each person knew each thing. The timeline is checked with the participants.
5. Learning interviews. One at a time, in a private place, on paid time, with accompaniment if the member wishes. The interviewer uses the cognitive interview method (Fisher & Geiselman, 1992): free recall first without interruption, then open questions in the order the member chooses, then specific questions, then a check. The question is always how it looked at the time, not why the member did not do something else. Interview notes are the facilitator’s and are never filed with personnel records. The guide is in the toolkit.
6. Local rationality. For each decision point the team writes what the person knew, what they were trying to do, what cues they had, what pressures they were under, and what made the action a reasonable one at the time (Woods & Cook, 1999; Dekker, 2014). Only then does it ask what the system could have given them that it did not. This guards against hindsight bias, the tendency to judge a decision by an outcome the decider could not see (Fischhoff, 1975)DOI.
7. Systems analysis. Contributing factors are classified against the work-system model (person, task, tools and equipment, organization, internal environment, external environment) (Holden et al., 2013)DOI and against the five recurring fireground factors (risk assessment, command, accountability, communications, procedures) (Fire Engineering, 2023). No factor is recorded as “human error” without the condition that produced it. The classification table is in the toolkit.
8. No-blame language. Reports identify people by role, never by name. They describe conditions, not character. They do not use “failed to,” “should have,” “ignored” or “complacent.” The facilitator checks the draft against the language rule before release.
9. Recommendations. Each is linked to a contributing factor, placed on the hierarchy of controls (NIOSH, n.d.-d), rated stronger, moderate or weaker by the rubric in the toolkit, given a priority from P1 (Imminent) to P4 (Improvement), a proposed owner and the due dates that priority sets, and sent to the Corrective Action Register. Reviews favour stronger actions and say so when only weaker ones are available.
10. Quality check. The sponsor checks the report against the quality standard in the toolkit (timeline complete; local rationality written for every decision point; every factor classified; every recommendation linked, rated and owned; language rule met; participants briefed first) before it is released.

Stage. S1 for steps 1 to 6 and 9; S2 for the full classification and the quality check measured.

The JOHSC, through a worker member, participates in the investigation of every event on the statutory list and every Tier 2 and Tier 3, as the applicable OHS statute provides (J05 Joint Committee and Health and Safety Representative). The member is on the team, not briefed afterward. The committee receives the triage log at each meeting, the Tier 2 and Tier 3 reports when released, and the register status quarterly. Any written recommendation the committee makes on the basis of a review is answered in writing within the statutory time. Where the committee and the department disagree on a recommendation, both positions are recorded in the register. Full workflow: 02 JOHSC Integration.

Stage. S1.

Requirement. Every recommendation from any tier, any JOHSC recommendation arising from a review, and any regulator order is logged in the Corrective Action Register and tracked to verified closure, which is the action-plan and validation step of the investigation standard (CSA Z1005:21, cl. 7) and the corrective-action clause of the management-system standard (ISO 45001:2018, cl. 10.2). The documented steps answer the accreditation criterion on addressing identified hazards (CPSE, 2019, criterion 11A.3).

The 30-day decision rule. Within 30 days of a report’s release, the fire chief (or the deputy chief to whom the decision is delegated) records for each recommendation one of: accepted, with owner and due date; accepted with modification, with the modification and the reason; deferred, with the reason, the interim controls, and a revisit date; rejected, with the reason. “No decision” is not an available status. The rule requires a documented rationale for accepting, rejecting or delaying each recommended action. Where the statutory JOHSC response clock is shorter than 30 days, the statutory clock governs (J05 Joint Committee and Health and Safety Representative). A hazard rated P2 (Serious) or higher does not wait for the report or the decision: it goes on the register when it is found, with the interim control and plan dates its priority sets.

Closure. An item is closed only when the owner has done the work, the verifier has confirmed it, and an effectiveness check has been recorded: did the change do what the recommendation intended? For a procedure change, that means a learning-team check-in with the crews who use it; for equipment, a field trial result; for training, a competency check. Items that fail the effectiveness check reopen.

Escalation. Overdue items escalate as the register’s scale sets for their priority: a P1 to the fire chief at once; a P2 to the deputy chief at the due date and the fire chief at 30 days overdue; a P3 (Moderate) to the deputy chief at 30 days overdue and the fire chief at 60 days. Open-item age is a standing management review input (ISO 45001:2018, cl. 9.3).

Evidence. The register; the decision records; the effectiveness checks.

Stage. S1 for the register and the decision rule; S2 for the effectiveness check measured.

The sharing element exists because posting a report does not by itself produce learning (Wildland Fire Lessons Learned Center, n.d.) and because the accreditation criterion asks that near-miss reporting be used to teach and share lessons (CPSE, 2019, criterion 11A.6). The department keeps its own near-miss database designed so that a provincial or national aggregator can take a feed from it later. The national near-miss system that ran from 2005 and reached several thousand reports (IAFC, n.d.) was, as of October 2026, no longer operating in its original form.

Stage. S1 for change notices and bulletins; S2 for the quarterly digest and training cases; S3 for external sharing.

Every year, before the management review, the program manager evaluates this program against: the reporting counts and trend by station and segment; triage timeliness; tier completion times; firewall integrity (the JOHSC’s check); recommendation decision timeliness; register age and effectiveness-check rate; bulletin timeliness; the culture baseline results for the six learning-system constructs (05 Safety Culture Baseline section 4); and the lessons the program itself has drawn. The evaluation is the department’s program evaluation for the accident investigation element under the responder safety standard (NFPA 1550, 2024, ch. 6 and annex H) and an input to Management Review (ISO 45001:2018, cl. 9.3). Every third year the external review examines it. Outline in the toolkit.

Stage. S2; external review S3.

This section is the heart of the note. The program exists to do these five things; everything else is the machinery.

Before the department writes or changes any reporting form, triage rule or review method, a learning team of six to ten members (career and volunteer, firefighters and company officers, with a JOHSC worker member present) describes how reporting and reviewing actually happen now (Conklin, 2012; Hollnagel, 2014): what gets reported and what does not, and why; what a member does the night of a bad call; who they tell; what happens to a report after it leaves the station; whether anyone hears back; what the hot debrief is really like; what people say about the last review the department ran; and what “near miss” means to them. The write-up names no one. The program is then designed to fit the route members already trust, not the one an office imagined. The same session is repeated at the 12-month evaluation.

Reporting and learning go right most of the time because of a few conditions: an officer who asks “anything we need to write up?” at the end of every call; a form that takes a minute; an HSO who answers within days; a chief who has visibly acted on a report; and a JOHSC worker member on the triage group whom members know by name. What erodes that capacity: a busy platoon with no time to debrief; a volunteer station where the officer is not seen for a week (NZFS, 2016b); a form that needs a login; a report that vanished; a review that took six months; a bulletin nobody read; a facilitator who has not run a review in two years; and any event, however old, where someone was disciplined after speaking up. The program protects capacity by setting clocks on acknowledgement, triage and review; by keeping two trained facilitators current through at least one review a year each; by measuring the time from report to feedback; and by making the chief’s rationale for every recommendation visible.

Errors, adaptations and workarounds found by a review are treated as information about the work system (Dekker, 2014). The learning firewall (element 5.5) applies to everything this program collects. Conduct concerns go to a separate accountability process run by different people, under the conduct policy and the collective agreement, and that process never receives learning interview content (Marx, 2001; Dekker, 2016). No discipline rests on learning information. Officers are trained to ask what made sense before asking who, and the review language rule makes the report say so.

A member who thinks an event should be reviewed says so to their officer, to the HSO, to a JOHSC member, or on the near-miss form, in any words; the request for a higher tier is recorded with the triage decision. A member who disagrees with a review’s findings says so to the facilitator before release and the disagreement is recorded in the report. A member who believes a review is being used against anyone tells the JOHSC worker co-chair, who raises it with the fire chief within five working days. The officer’s expected reply to any of these is: acknowledge, record, pass it on, and tell the member what happened. Leaders show it is safe by reporting their own near misses and good catches by name in the bulletin, by sitting in learning-team sessions without speaking first, and by publishing the rationale when they reject a recommendation (Edmondson, 1999DOI; Conklin, 2019). The indicator is the count of near-miss and good-catch reports per working fire, by station, which should rise in the first year, and the survey item “I can raise a concern with my officer and it will be dealt with.”

This program feeds all three tiers; it is the program the tiers belong to. The two questions the program asks of itself at every annual evaluation: “Which events this year did not get reported until later, and why?” and “Which recommendation did we close that did not actually change anything?” For the full method, worked examples and the decision flow, see Manual 02 - The Three Tiers and Event Triage; for the policy, see Manual 08 - Model Policy and the department’s own Policy - Incident, Injury, Exposure and Near-Miss Reporting and Investigation.

Certificates of completion only; the department does not “certify” investigators.

This program has little equipment. What it has:

• SWP - Exposure Reporting
• The hot debrief and AAR procedure (the four-question sheet, Form - After-Action Review (Tier 1))
• Scene preservation after a serious event (in the statutory reporting procedure, element 5.2; see J08 Incident Reporting and Investigation)

All records are written for freedom-of-information disclosure: roles, not names; conditions, not character. Exposure records are kept as the accreditation criterion and the responder safety standard require (CPSE, 2019, criterion 11A.10; NFPA 1550, 2024, ch. 15).

Quarterly indicator worksheet in the toolkit.

Loop test. A firefighter at a volunteer station reports on Monday night, by QR code, that a crew went to the second floor of a working fire without the accountability board being set up. By Tuesday noon the Triage Group has it as a Tier 2 (significant operational failure); the member has an acknowledgement; the sponsor and facilitator are named by Thursday. The PIA reports within 30 days with contributing factors classified and three recommendations rated. The deputy chief decides each within 30 days and records why. A change notice goes to every station within 15 days of the decision, and a learning bulletin follows. The member who reported hears, from the HSO, what was found and what changed, and the item closes only after a learning-team check-in at that station confirms the board goes up now.

Checklist: Checklist - Incident Reporting, Investigation and Learning. The five questions an auditor asks first:

1. Show me the last ten near-miss reports and, for each, the date the reporter was told what happened.
2. Show me the triage record for the last three events above Tier 1 and the trigger that set each tier.
3. Show me the last Tier 2 report: where is the local rationality for each decision point, and which recommendation is rated stronger?
4. Show me the chief’s recorded decision, with rationale, on every recommendation from the last two reports, and the effectiveness check on every closed item.
5. Ask a firefighter at a volunteer station what happens when they report a near miss. Then ask their officer. Do the answers match the policy?

This program answers the evidence in 05 Design Principles and Lessons Learned directly. A public inquiry into a Canadian structural collapse response found after-action reports absent across agencies, some apparently to avoid criticism, and recommended they be mandatory (Firefighting in Canada, 2014a); Tier 1 is mandatory and short for that reason. Research with volunteer brigades found non-reporting driven by fear of blame, not wanting to inform on mates, paperwork, not knowing what a near miss is, and never hearing back (NZFS, 2016b); the 60-second path, the posted definition, the firewall and the feedback clocks each answer one of those. The staff of the most mature fire-service learning system observe that posting a report does not produce learning (Wildland Fire Lessons Learned Center, n.d.); the change notice, the bulletin and the effectiveness check exist so that a report becomes a change in how work is done. A provincial regulator has routed its findings to the fire school and the provincial operations guide (Gouvernement du Québec, 2022); this program’s external sharing element does the same from the department’s side. The five recurring fireground factors (Fire Engineering, 2023) are the second axis of the contributing-factor classification, so the department can see over time whether its events cluster on risk assessment, command, accountability, communications or procedures. And the stability of the NIOSH recommendation list over 16 years (Hard et al., 2019)DOI is the reason the program measures whether a closed recommendation actually changed anything.

Stage 1 minimum (S1). The policy adopted with the firewall clause agreed by the bargaining agents and the volunteer association; the statutory reporting procedure with the call log; the near-miss path (QR, paper, phone) and the posted definition; the injury and exposure form; a Triage Group with backups and the triage record; Tier 1 AAR on every working fire; two trained Tier 2 facilitators; the Corrective Action Register with the 30-day decision rule; change notices; the JOHSC on every serious-event investigation.

Cost. Two facilitators trained (about two days each plus a supervised first review); half a day of officer training per company officer; printing and a QR code; about four hours a month of HSO time for acknowledgements and the register in a mid-size department, more in the first year; a retainer or standing agreement for an independent reviewer so that Tier 3 is arranged in advance, not found after the event. A volunteer department can run everything in this program with one trained facilitator shared with a neighbouring department under a written mutual arrangement, which also gives each the independence the other lacks.

This chapter relies on the following legal requirements, which differ by province and territory. Check your jurisdiction’s section on each topic page before adopting it.

Full form: Form - Near-Miss and Good-Catch Report. The six fields of the 60-second version: what happened (free text); where and when; what could have happened; what stopped it or what would have; what you think would prevent it; how to reach you if you want to hear back (optional). The firewall statement is printed on the form.

Full form: Form - Injury and Incident Report. Member section: who (role, station, status career or volunteer); when and where; activity at the time; what happened; body part and nature of injury, or substance and route of exposure, or equipment and failure mode; PPE worn; witnesses (roles); first aid or treatment; whether the member is fit to continue. Supervisor section: immediate actions; scene preserved; statutory list checked; reported to HSO and district chief (times); initial view of conditions; referred to triage. HSO section: statutory reports made (times); WCB notice; triage tier; register items.

Full form: Form - Event Triage Record. Fields: event reference; date and time reported; date and time triaged; group members (roles); triggers met, by tier; tier set; escalation requested by anyone (yes, by whom by role, outcome); de-escalation (reasons, unanimous); statutory reporting status; support offered; preservation status; sponsor and facilitator named (date); next review date.

Full form: Form - After-Action Review (Tier 1). The four questions, after the military after-action review method (The Human Factor, 2026a): What was supposed to happen? What actually happened? Why was there a difference, and what went well that we should keep? What will we do differently, by whom, by when? Plus: anything for the register; anything that needs a higher tier; who was present (roles).

Post-Incident Analysis: Terms of Reference
Event reference and short description:
Tier set by triage on (date), trigger(s):
Sponsor (role):                           Lead facilitator (role, trained on):
Team: JOHSC worker member (role); operational officer not at the event (role); subject matter adviser(s) (role)
Scope: what is in (the incident from dispatch to demobilization; the training evolution; the station task); what is out (any conduct question, which goes to the separate process; any statutory process, which runs alongside)
Questions the review will answer:
  1. What happened, in sequence, and how did it look to the people in it at the time?
  2. What conditions shaped their choices (task, tools, organization, environment, external)?
  3. What went well, and what capacity did the crew draw on?
  4. What would make it go better next time, at the strongest level of control available?
Access: records (list), people (list by role), scene, equipment set aside
Timeline: terms of reference signed (within 72 hours of triage); interviews complete by; draft to sponsor by; report released by (30 days; 60 maximum)
Learning firewall statement: attached and given to every participant before interview
Support: peer support offered to all participants on (date), independent of participation
Statutory interfaces running alongside: regulator notice (date); WCB (date); JOHSC notified (date); counsel liaison (role)
Report audience and distribution: participants first; then sponsor, fire chief, JOHSC; bulletin after release
Signed: sponsor (role, date); lead facilitator (role, date)

Full outline: Form - Post-Incident Analysis Report Outline (Tier 2). Sections: summary (one page); terms of reference; method and sources; the event as experienced (narrative from the participants’ view); timeline; what went well; decision points with local rationality; contributing factors classified (table); recommendations linked, placed on the hierarchy, rated, with proposed owner and date; disagreements recorded; support provided; distribution and briefing record; appendices (de-identified).

The guide applies the cognitive interview (Fisher & Geiselman, 1992) and the critical decision method (Klein et al., 1989)DOI to a fire service setting.

Before. Choose a private room away from the station floor. Offer accompaniment (peer, union or association representative). Give the firewall statement in writing and read it aloud. Explain the purpose: to understand how it looked at the time so the department can change the conditions. Say that notes are the facilitator’s and will be destroyed when the report is accepted. Ask permission to take notes. One interviewer leads; a second may take notes.

Opening. “Take me through it from when you first heard the call, in whatever order it comes to you. I will not interrupt.”

Free recall. Let the member talk until they stop. Then: “Is there anything else?” Wait. Then again.

Open questions, in the member’s order. “What did you see at that point?” “What were you trying to do?” “What did you expect to happen next?” “What was the radio like?” “Who else knew what you knew?” “What were you worried about?” “What would have told you it was going wrong?”

Specific questions. Only after the open phase: times, positions, equipment, what was said. Use the timeline. Check differences without challenging: “Our records show the PAR at 14:22; how does that fit with what you remember?”

What went well. “What did you or the crew do that kept it from being worse?” “What do you usually do on a job like this that worked here?”

Conditions. “If you could change one thing about how we set this up, what would it be?”

Close. Summarize what you heard and ask what you got wrong. Say what happens next and when they will hear. Repeat the support offer. Thank them.

Rules. No “why didn’t you.” No leading questions. No interviewing two people together. No recording without consent and the bargaining agent’s agreement. Stop the interview if the member becomes distressed and reconnect them with support; resume another day. If a conduct concern surfaces, stop, do not probe it, and tell the sponsor only that a concern exists.

Classify every factor on both axes: the work-system element (Holden et al., 2013)DOI and the recurring fireground factor (Fire Engineering, 2023), if any. Enter the factor in one work-system row and tick the fireground factor it relates to.

Rule: no factor is entered as “human error,” “complacency,” “failed to” or “lost situational awareness” without the condition in another row that produced it.

The rubric adapts the hierarchy of controls (NIOSH, n.d.-d) and the healthcare hierarchy of intervention effectiveness (ISMP Canada, n.d.).

A review that can offer only weaker recommendations says so and names the stronger action it could not recommend and why. The register records the rating, and the annual evaluation reports the share of stronger actions among those adopted.

LEARNING BULLETIN {{Bulletin number}}                                    {{Department}} · date
What happened (as the crew experienced it, de-identified, 150 words)
What we found (the conditions, not the people: three to five bullets from the classification table)
What went well (one thing the crew did that kept it from being worse)
What the department is doing (each recommendation: action, rating, owner by role, date, status)
What a crew can do tonight (one or two things, practical)
Where to read more (the report reference; the related procedure; the training case)
Say thank you: this bulletin exists because someone reported. Report yours: {{QR code / phone}}.
Issued by: HSO (role). Reviewed by: participants on (date); JOHSC on (date).

1. Summary for the chief and council (one page): events by type and tier; what changed because of them; the three lessons of the year; what the department asks for.
2. Reporting: counts per 100 calls and per working fire by station and segment; trend against baseline; acknowledgement and feedback times; the learning-team findings on reporting.
3. Triage: events triaged, timeliness, escalations and de-escalations with reasons, disagreements.
4. Reviews: Tier 1 completion; Tier 2 count, timeliness, quality checks; Tier 3 count and status; facilitator currency.
5. Firewall integrity: the JOHSC’s check and its result; any concern raised and how it was resolved.
6. Recommendations: count by rating; decisions within 30 days; register age; effectiveness checks; items reopened.
7. Sharing: change notices, bulletins, training cases, external sharing.
8. Culture: the six learning-system construct bands from 05 Safety Culture Baseline and the psychological-safety indicators.
9. Statutory: notices made and their timeliness; regulator contacts; JOHSC recommendations and responses.
10. The program’s own lessons: what did not work in the program this year and what will change.
11. Resources: facilitator time, training, the Tier 3 arrangement, cost.
12. Decisions requested of the management review.

You are being asked to take part in a learning review. Its purpose is to understand how the event looked to the people in it and what conditions shaped what happened, so the department can change those conditions. What you say here is used for learning only. It will not be used as the basis for discipline, performance management or any adverse action, and it will not be placed in your personnel file. Notes taken are the facilitator's and are destroyed when the report is accepted. The report identifies people by role, not by name. If a conduct concern (intentional harm, impairment, reckless disregard, a criminal act) arises, the review will stop and the concern will be referred to a separate process that does not receive this interview; you will be told if that happens. Statutory processes (regulator, workers' compensation, coroner, police) run separately and may require records by law. You may bring a peer, union or association representative. You may take a break or stop at any time. Peer support is available to you whether or not you take part. Signed: lead facilitator (role, date). Received: participant (initials, date).

Institute for Safe Medication Practices Canada [ISMP Canada]. (n.d.). Hierarchy of effectiveness [Web page]. https://ismpcanada.ca/resource/hierarchy-of-effectiveness/

International Association of Fire Chiefs [IAFC]. (n.d.). Near-miss reporting: 5,000 reports and counting. https://iafc.org/iCHIEFS/iCHIEFS-article/near-miss-reporting-5-000-reports-and-counting

National Institute for Occupational Safety and Health [NIOSH]. (n.d.-d). Hierarchy of controls [Web page]. Centers for Disease Control and Prevention. https://www.cdc.gov/niosh/hierarchy-of-controls/about/index.html

Wildland Fire Lessons Learned Center. (n.d.). About us [Web page]. https://lessons.wildfire.gov/about-us