Incident investigation and root cause analysis: a practical guide

When something goes wrong at work, whether an injury, a near miss, a spill or a quality failure, the investigation decides whether it happens again. A good investigation finds out why it happened and fixes the system. A weak one stops at “operator error” and closes the file.

This guide sets out a practical approach to incident investigation and root cause analysis that works for health and safety, environmental and quality incidents.

First: make safe, preserve, report

• Make the scene safe and look after anyone injured.
• Preserve evidence: keep the area, equipment, documents and data as they were, and take photographs before anything is moved.
• Report as the law requires. In Malaysia, accidents causing death or serious injury, dangerous occurrences and occupational diseases must be notified to the Department of Occupational Safety and Health (DOSH) under the Occupational Safety and Health (Notification of Accident, Dangerous Occurrence, Occupational Poisoning and Occupational Disease) Regulations 2004 (NADOPOD). Check the regulations and DOSH guidance for what to report, how and when. Separate reporting duties may apply, for example to SOCSO, your insurer or your client.

Step 1: Gather the facts

• Interview the people involved and witnesses separately and as soon as possible, in a fact-finding tone, not a disciplinary one.
• Collect documents: permits, procedures, training records, maintenance records, shift logs, CCTV and system data.
• Build a timeline of what happened, minute by minute where it matters.

Step 2: Analyse why it happened

Most incidents have several causes at different levels:

• Immediate causes: the acts and conditions present at the time, such as a guard removed or a valve left open.
• Underlying causes: why those acts and conditions existed, such as an unclear procedure, time pressure or a missing inspection.
• Root causes: the management system failures behind them, such as how work is planned, how risks are assessed, how contractors are managed or how changes are controlled.

Useful techniques

• Five whys: keep asking “why” until you reach a cause that management can fix. Simple and quick, but it can follow a single line and miss other causes.
• Cause-and-effect (fishbone) diagram: groups possible causes under headings such as people, methods, equipment, materials, environment and management, so nothing is overlooked.
• Barrier analysis: asks which controls should have prevented the incident, and why each one failed or was missing.

For serious incidents, a structured method such as ICAM or a similar systematic approach is worth the extra effort, ideally led by someone independent of the area involved.

Step 3: Agree corrective actions

• Address each root cause, not only the immediate one.
• Prefer stronger controls: eliminate or engineer out the hazard before relying on procedures, training or personal protective equipment.
• Give each action an owner and a due date, and decide how you will check that it worked.
• Share the lessons with other sites and teams doing similar work.

Step 4: Close the loop

ISO 45001, ISO 9001 and ISO 14001 all require organisations to react to incidents and nonconformities, determine their causes, take corrective action and review its effectiveness. Track actions to completion, look for trends across incidents and near misses, and report them to management.

Common pitfalls

• Blaming the individual. “Human error” is a starting point for the investigation, not a conclusion.
• Investigating too late, after evidence has gone and memories have faded.
• Actions that only say “retrain” or “remind staff”, which rarely prevent a repeat.
• Ignoring near misses, which are free lessons about the same weaknesses that cause serious incidents.
• Investigating your own area. A degree of independence makes findings more credible with management, clients, insurers and regulators.

Next step

An independent investigation, or support for your own investigators, brings structure and objectivity to serious incidents. Training your supervisors in root cause analysis builds the same discipline into everyday incident reporting.

This article is general information, not legal or professional advice. Requirements can change; check the current law and guidance, or speak to us about your circumstances.