The bowtie method is a proactive risk analysis tool, which is used to understand risks and to prevent incidents from occurring. In one of our previous blogs, Human Factors were discussed in bowtie risk modeling. There are different ways to include Human Factors, for example in the threat or in an escalation factor (or degradation factor). We should also not forget to understand the strengths of the human act in a barrier system. There are many different best practices on how to include Human Factors in bowties and this also applies to incident analysis methodologies.
Despite best effort in risk assessments, incidents can still occur. In order to learn from incidents and review the bowtie, various incident analysis methods are designed and they all include Human Factors in a different way.
IncidentXP uses 5 different methods:
- Barrier Failure Analysis (BFA)
- Barrier-based Systematic Cause Analysis Technique (BSCAT)
- Tripod Beta
- Root Cause Analysis (RCA)
- Root Cause Analysis TOP-SET
In this blog, we’ll take you through the guidelines of including Human Factor when using these methods.
Root Cause Analysis (RCA)
The reason to start this blog with the Root Cause Analysis (RCA) method is that this is a commonly used method amongst our clients and partners. RCA in IncidentXP is flexible to use since we only work with event boxes. The pitfall here is that there is no distinction between events, barrier failures, immediate-, secondary- and tertiary causes. It is therefore difficult to identify Human Factors in the incident scenarios.
The way Human Factors can be taken into account in RCA is by categorizing the event boxes, like in figure 1. By doing it this way, it is clear what events were human-related.
Figure 1: Human Factor in Root Cause Analysis (Note: Underlying (secondary) causes are not visualized in this diagram).
Root Cause Analysis (RCA) TOP-SET
The RCA TOP-SET method works quite the same as the RCA method discussed previously. In addition, this method provides more structure and uses the TOP-SET method in the whole incident investigation process, to eventually build an RCA diagram.
TOP-SET stands for Technology, Organization, People, Similar Events, Environment, and Time Sequence (and History of Events). These indicators are constantly used during the incident investigation and analysis process. One of the guidelines, which includes Human Factors, is to define the root cause in the Organization or People category, instead of defining the root cause as a technical failure. (The Kelvin Consultants Ltd, 2018).
Barrier-based Systematic Cause Analysis Technique (BSCAT)
The BSCAT™ method is the barrier based extension to DNV’s SCAT method. It is used to identify barrier failures and to link the causations to the SCAT chart. The SCAT chart is a standardized list of immediate causes, basic causes, and management system factors. Due to using a standardized list, incident analyses are suitable for aggregation, leading to more insight into the weak areas of your safety management system.
The BSCAT method is a good handhold to include Human Factors in incident analysis. It is included in the immediate causes (substandard acts) and the basic causes (personal factors), see example below. The management system factors are set up to identify the root cause of the barrier failure and to understand where actions of improvements need to be made.
Figure 2: Example BSCAT with causation.
Barrier Failure Analysis (BFA)
The Barrier Failure Analysis (BFA) makes a distinction between events, barrier failures, and causations. The structure of the analysis is almost the same as in BSCAT™, except for the standardized list of causations. Basically, the BFA is BSCAT without the SCAT categories and indicators.
The advantage of using BFA is that other human factor or human error theories can be used within this method. See for example, the Human Factors Analysis and Classification System (HFACS) (Shappell & Wiegmann, 2000). This framework is designed to use in the aviation industry and can be implemented in BFA and linked to the causation paths. This is also the case for any other theory which uses lists of causations.
It is useful to use these kinds of frameworks to identify why barriers fail, but also to see trends in barrier failures. Each of these frameworks has a different way to approach Human Factors.
Tripod Beta has excellent guidelines to include human factor in incident analyses. The method uses, amongst others, the Generic Error-Modelling System (Rasmussen, 1987) in their causation pathways. Compared to BFA and BSCAT, Tripod has an extensive and flexible causation path. In Tripod, failed barriers can only have one immediate cause, but multiple preconditions and underlying causes.
The immediate cause is always related to the act (or non-act) of a person (Stichting Tripod Foundation, 2014). In the example below (figure 3), you see the immediate cause is directly related to barrier failure. The method works back to the underlying causes to find out what has caused the person to make an error.
Tripod Beta is commonly used for complex incidents and gives a good understanding of Human factors that played a role in the incident. However, Tripod Beta might not be the best method to use for small incidents and near misses, because it will make a relatively simple incident unnecessarily complex.
Figure 3: Tripod Beta causation path.
In summary, the methods available in IncidentXP all offer a way to dig into the deeper underlying causes of an incident. Most techniques use barrier-based thinking as a central way to do this. Identifying which Human Factors were part of the Incident is often only the first step (immediate cause/primary cause) of the incident analysis. The next step is to go further down the causation pathway and find the underlying causes. Once the underlying cause is identified, specific actions of improvement can be made to eventually prevent incidents in the future.
However, it is also good to understand the Human Factors as a strength in incident analysis. Think about the barriers that include human acts that prevented a major consequence.
Human Factors can be taken into account in various ways in IncidentXP. It is good to notice that there are multiple methods, which all has its strengths and weaknesses. From CGE’s point of view, it is important that the method helps to understand incident scenarios rather than following all the guidelines. We are very curious to hear your thoughts on this subject, so feel free to leave a comment below or to contact us in case of any practical questions.
Learn about Human Factors from Prof. Dr. Jop Groeneweg
In this blog, we’ve shown you how to visualize Human Factors in IncidentXP, but to find and name the specific Human Factors is a wholly different skill. For this, we trust our partners in the academic world such as Prof. dr. Jop Groeneweg from Leiden University.
Curious to find out what Prof. Dr. Jop Groeneweg has to say about Human Factors? Sign up for the Risk Management Master Class where Prof. Dr. Jop shares his insights during the Safety Culture & Human Factor course.
Shappell, S. A., & Wiegmann, D. A. (2000). The Human factorsHuman factors Analysis and Classification System-HFACS. U.S. Department of Transportation. Washington: Federal Aviation Authority.
Stichting Tripod Foundation. (2014). Tripod Beta. London: Energy Institute.
The Kelvin Consultants Ltd. (2018, April 3). Root Cause Analysis. Retrieved from Kelvin TOP-SET professional incident investigation: http://www.kelvintopset.com/about/root-cause-analysis
Rasmussen J. (1987). Cognitive control and human error mechanisms. In Rasmussen J., Duncan K., & Leplat J., New Technology and Human Error. Chichester: John Wiley & Sons, Ltd.