2.4 Incident Investigation & Root Cause Analysis (RCA)

Incident Investigation & Root Cause Analysis (RCA)

The Purpose of Investigation

When a safety incident occurs—whether it is a near miss, property damage, or a severe injury—a thorough investigation is mandatory. The absolute, uncompromising goal of an incident investigation is fact-finding, not fault-finding. The objective is to identify exactly what happened, why it happened, and what corrective actions are necessary to prevent a recurrence.

If the organizational culture focuses on blaming individuals (e.g., concluding an investigation with "worker was careless"), reporting will cease, near misses will be hidden, and the true systemic failures will remain uncorrected, guaranteeing future accidents.

The Investigation Process

An effective incident investigation follows a structured sequence:

1. Secure the Scene

Immediately after ensuring injured personnel receive medical care, the scene must be secured. Barricade the area to prevent alteration of evidence. Only authorized investigators should enter. Equipment should remain exactly as it was at the time of the incident (unless moving it is necessary for rescue or to prevent further imminent danger).

2. Gather Evidence (The 4 P's)

Evidence degrades rapidly. The investigation team must collect data across four categories:

• People: Identify witnesses and conduct interviews as soon as possible. Memories fade and become influenced by talking with others. Interview witnesses separately. Ask open-ended questions.
• Parts: Examine the physical equipment, tools, and machinery involved. Look for broken components, bypassed guards, or signs of wear.
• Position: Document the physical environment. Take extensive photographs from multiple angles. Create sketches. Note environmental conditions (lighting, noise, weather, housekeeping).
• Paper: Secure relevant documentation. This includes JSAs, training records, maintenance logs, standard operating procedures, and recent inspection reports.

3. Analyze the Sequence of Events

Reconstruct a timeline leading up to the incident. Determine the exact sequence of actions and conditions. This timeline must be based on facts established by the evidence, not assumptions.

4. Determine Causes (Direct vs. Root Cause)

An incident is rarely the result of a single failure. Investigators must peel back the layers to identify different levels of causation:

• Direct Cause: The immediate event or condition that caused the injury or damage. (e.g., Worker slipped on an oil spill).
• Contributing Causes: Factors that contributed to the incident but would not have caused it alone. (e.g., Worker was not wearing slip-resistant shoes; lighting in the area was poor).
• Root Cause(s): The underlying systemic, management, or organizational failures that allowed the direct and contributing causes to exist. (e.g., The preventive maintenance program for the machine leaking oil was underfunded and behind schedule; safety footwear policies were not enforced by supervision).

Fixing the direct cause (cleaning up the oil) solves the immediate problem. Fixing the root cause (repairing the machine and overhauling the maintenance schedule) prevents the problem from ever happening again.

Root Cause Analysis (RCA) Techniques

Several analytical tools assist in moving past direct causes to uncover root causes:

The '5 Whys' Technique

This is a simple, iterative interrogative technique used to explore cause-and-effect relationships. You ask "Why?" repeatedly until you reach a systemic failure.

• Event: Worker suffered a laceration to the hand while operating a bandsaw.
• Why? His hand came into contact with the moving blade. (Direct Cause)
• Why? The adjustable blade guard was positioned too high. (Contributing Cause)
• Why? The worker did not adjust the guard before making the cut.
• Why? The worker was hurrying to meet a production quota and bypassed the adjustment step.
• Why? Supervisors are incentivizing production speed over safety protocols, and no refresher training on machine guarding has been conducted in two years. (Root Cause)

Fishbone Diagram (Ishikawa Diagram)

Also known as a cause-and-effect diagram, this visual tool helps brainstorm and categorize potential causes of a problem. The "head" of the fish is the incident. The "bones" branching off represent major categories of causes, typically:

• Man / People: Lack of training, fatigue, human error.
• Machine / Equipment: Malfunction, lack of maintenance, poor design.
• Method / Process: Inadequate procedures, lack of JSAs.
• Material: Defective raw materials, hazardous chemicals.
• Environment: Poor lighting, extreme temperatures, noise.

Investigators brainstorm specific causes under each category to visualize the complex interplay of factors.

Fault Tree Analysis (FTA)

FTA is a top-down, deductive failure analysis using Boolean logic (AND / OR gates) to analyze the various ways a specific undesired event (the "top event") can occur. It is highly structured and often used in complex engineering and process safety incidents.

Implementing Corrective Actions

The final step is generating a Corrective and Preventive Action (CAPA) plan. These actions must target the root causes. Every corrective action must be Specific, Measurable, Assignable (to a specific person), Realistic, and Time-bound (SMART). The OHST plays a crucial role in tracking these actions to ensure they are completed and validating that they actually resolved the systemic issue without introducing new hazards.