Mark Paradies and
Benna Hughes discuss
Incident Investigation

Incident Investigation

Incident Investigation is a systematic process to find the root causes of problems and develop effective solutions to improve performance. This article includes:

  • A definition of an incident,
  • The best way to investigate simple incidents,
  • How to investigate major accidents,
  • Best practices for collecting incident information (including interviewing),
  • Developing the incident’s sequence of events,
  • Identifying the incident’s Causal Factors,
  • How to find an incident’s root causes and Generic Causes,
  • Developing corrective actions to prevent repeat incidents,
  • Getting management to understand what happened and to approve the resources needed to implement corrective actions to prevent future incidents, and
  • Measuring the effectiveness of your incident investigations.

Definition of an Incident

An incident is:

The worst consequence that
happened in a sequence of events.

The sequence of events describes what happened in a safety, quality, equipment reliability, production, or environmental problem. Thus an incident is the worst safety, quality, equipment reliability, production, or environmental issue in the sequence of events.

It is possible to have multiple incidents in a single chain of events. For example, you could have injuries or fatalities, equipment damage, and an environmental release as part of a sequence of events that includes a chemical plant explosion.

Another type of incident to investigate is a precursor incident. A precursor incident is:

Precursor Incident
Minor incidents that could have been a major accident
if one or two more Safeguards would have failed.

Precursor incidents may be called close calls or near-misses.

For a more in-depth discussion about precursor incidents, CLICK HERE.

Now that we understand the definition of an incident, we must find the best way to perform an incident investigation.

Simple Incident Investigations

Most incidents are low-to-moderate risk incidents, precursor incidents, or close calls (near-misses).

The best way to investigate any incident, even minor incidents that are still worth investigating is to use a systematic process. That systematic process should avoid common problems (placing blame and cognitive biases) and promote the use of investigation best practices.

Even simple investigations have a minimum set of simple incident investigation best practices to succeed. What are these best practices? CLICK HERE for an article that explains them.

Is that minimum set of best practices just too much? Then you have two choices:

  • If the incident isn’t worth investigating … DON’T investigate it!
  • Try the fastest cause analysis tool – Spin-A-Cause™! (OK – that’s a joke.)

Major Accident Investigations

What is the difference between a simple investigation of a precursor incident and the investigation of a major accident? That could be the subject of a whole book! But let’s start with the fact that a major accident investigation will be:

  • much more complex,
  • will probably have a team performing the investigation, and
  • will have more management and regulatory interest (and perhaps even a separate regulatory or prosecutorial investigation).

Thus a major accident investigation process will have to be more robust to accomplish the goals of the investigation.

Some other differences may include the following:

  • Performing the investigation under the direction of an attorney to maintain legal privilege.
  • A more highly trained investigation leader/facilitator and more highly trained investigators.
  • Assistance from specially trained consultants (for example, equipment or human factors experts).
  • Optional information collection, interviewing, and root cause analysis tools that aren’t typically applied in simple incident investigations.
  • More reviews and approvals for the final investigation findings/report.

For a major investigation process combined with industry-leading root cause analysis tools, CLICK HERE to order a book about performing major investigations.

Collecting Evidence and Interviewing

An investigator must collect accurate information to perform an incident investigation, find root causes, and develop effective corrective actions. We categorize this evidence as 3 P and an R. That is:

Here are some tips for collecting better evidence:

For more information on interviewing and evidence collection training, CLICK HERE.

Finding the Incident’s Causal Factors, Root Causes, and Generic Causes

Once you have collected the investigation informant (evidence) and displayed it in a sequence of events, you are ready to find the incident’s Causal Factors, root causes, and Generic Causes. How do you find root causes? See THIS PAGE.

Developing Corrective Actions

Investigators can fall into traps when developing corrective actions to prevent repeat incidents. What kind of traps? Here are two examples:

  • You are using the easiest corrective actions rather than the most effective corrective actions. See THIS ARTICLE and learn more about using the hierarchy of controls to develop effective corrective actions.
  • THIS VIDEO provides ideas from Dave Janney and Benna Hughes to help investigators understand and avoid corrective action traps (especially an idea about corrective actions that sound good but don’t work).

Management Approval of Incident Investigation Reports

What does management need from your report (or presentation) to be able to approve your incident investigation? Here are five critical sections for every incident investigation report:

  • What happened? They need to understand the sequence of events.
  • Why did it happen? They need to understand how the Causal Factors, root causes, and Generic Causes fit into the sequence of events.
  • How can we fix the Causal Factors, root causes, and Generic Causes? They need to see a clear connection between the corrective actions and the things they are preventing and how that will stop a repeat incident.
  • What resources are required for corrective action implementation? Who will be responsible for implementing the corrective actions, and when can they be implemented if the resources are provided? Management should evaluate the schedule to determine if it is adequate for the risk profile of the incident.

Measuring the Effectiveness of Your Incident Investigations

The best incident investigation performance measure (performance indicator) is your facility’s rate of repeat incidents.

What is a repeat incident? Consider these examples:

  • Should the investigation and corrective actions for the Challenger Space Shuttle accident have prevented the Columbia Space Shuttle accident by improving flight safety at NASA?
  • Should the BP Texas City fire and explosion accident investigation have prevented the BP Deepwater Horizon accident by improving process safety at BP?

Review ten years of your facility’s incident history, starting with the oldest incidents. As you progress to newer incidents, do you see repeat incidents, Causal Factors, or root causes?

Generic Causes are, by definition, repetitive. Therefore, every incident with a Generic Cause is probably a repeat incident.

What is the repeat rate for your facility? Do you have 80% repeats? 10%? 0.1%?

We hope this incident investigation guidance has helped you develop ideas to improve your safety, quality, equipment reliability, operational excellence, and human performance investigations. If you want our help analyzing and improving your incident investigation program, please contact us by CLICKING HERE or calling 865-539-2139. Or, if you want to progress from a reactive-based incident investigation program to a proactive performance improvement program using state-of-the-art root cause analysis tools, once again, please contact us.

To receive a FREE weekly newsletter about incident investigations, root cause analysis, and performance improvement, register using the form below.