Incidents causality model

A specific example of a causal model is the root cause tree described in Section 6.8.4 and Figure 6.8. This is a very elaborate model which includes several levels of detail for both equipment and human causes of incidents. The root causes tree is a generic causal model, and may require tailoring for application to specific plants and processes (e.g., in the offshore sector) where other error causes may need to be considered. 

The causality model used in accident or incident analysis determines what we look for, how we go about looking for facts, and what we see as relevant. In our experience using STAMP-based accident analysis, we find that even if we use only the information presented in an existing accident report, we come up with a very different view of the accident and its causes. 

This chapter addresses the need for safety professionals to adopt an incident causation model, a thought process based on a sound understanding of the hazards-related incident phenomenon and which, when applied, identifies the reality of the causal factors in the incident process. 

Safety professionals apply differing and contradictory incident causation models, and the work of some of them is misdirected and ineffective. Professional safety practice requires that the advice given to avoid, eliminate, or control hazards be based on a sound incident causation model, a thought process, so that, through the application of that model, the desired risk reduction is attained. That will not occur if the causation model used does not require identifying the actual causal factors. 

This is a serious subject. For incident causal factors that are actions or inactions of individuals, their so-called errors may be programmed into the work system created for them. And a causation model has to address the programming sources. 

First, which incident causation model is to apply must be determined. All that follows in this outline regarding causal factors will relate to that model. 

Application of an incident causation model that properly balances causal factors deriving from less than adequate policies, standards, or procedures that impact on design management, operations management, and task performance... 

In addition to the proactive uses of the SRK model described in the two previous sections, it can also be employed retrospectively as a means of identifying the underlying causes of incidents attributed to human error. This is a particularly useful application, since causal analyses can be used to identify recurrent vmderlying problems which may be responsible for errors which at a surface level are very different. It has already been indicated in Section 2.4.1 that the same observable error can arise from a variety of alternative causes. In this section it will be shown how several of the concepts discussed up to this point can be combined to provide a powerful analytical framework that can be used to identify the root causes of incidents. 

Step 4. Every element in such a tree will be classified according to the chosen human behaviour model, or at least every "root cause" (the end points of the tree) will be. In this way the fact that any incident usually has multiple causes is fully recognised. Each near miss report is analysed to produce a set of classifications of causal elements instead of the... 

If several safety professionals investigate a given hazards-related incident, they should identify the same causal factors, with minimum variation. That is unlikely if the thought processes they use have greatly different foundations. At least 25 causation models have been published. Since many of them conflict, all of them cannot be valid. A review of some of them is followed by a discussion of principles that should be contained in a causation model. 

As used in this treatise, causation means the act or agency of causing or producing an effect. Causal factors include all of the elements — the events, the characteristics of things, and the actions or inactions of persons— that contribute to the incident process. A model is to represent the theoretical ideal for the process through which hazards-related incidents occur, a process that requires determining when the phenomenon begins and ends. 

It is a necessity that the advice given by safety professionals be based on an understanding of the reality of causal factors and actually serves to attain a state for which the risks are judged to be acceptable. I will develop a systemic causation model for hazards-related occupational incidents that represents the thoughts set forth in this chapter. 

Most causation models have minimized less than adequate design and engineering concepts and outcomes as a source of causal factors for hazards-related incidents, with one significant exception. That exception is MORT—the management oversight and risk tree. Concepts on which MORT is based have influenced my thinking greatly, and I am indebted to all who worked on the creation and betterment of MORT. 

To promote an understanding of this systemic, causation model, meanings are given here of the terms incidents, hazards, HAZRINS, model, causation, causal, factors, and causal factors. 

One last comment on the model—it relates not to causality but rather to increases and decreases in incidence of myocardial lesions. It seems that with high fat diets lesions will appear, irrespective of the oil type. Furthermore, 34% of the variation observed by Trenholm et a/. (1979) arises from differences among experiments [although analyses of covariance suggest that a considerable proportion of this variation (about 25%) can be attributed to differences in dietary fatty acid levels across experiments]. 

Incident analysis or development of preventative measures is mainly done based on a particular analysis model, which is used for obtaining a comprehensive picmre of incidents and for bringing out the relevance of causality. In this section, two principal models are introduced. 

In Chapter 3, Serious Injury Prevention, an outline for such a study was presented under the heading Proposing a Study of Serious Injuries. Such a study will not be time-consuming since the data to be collected and analyzed should already exist or can be obtained easily. To assist in such a study, two addenda are provided at the conclusion of this chapter. Both are reprinted Ifom the third edition of On The Practice Of Safety Addendum A, A Systemic Causation Model for Hazards-Related Incidents, and Addendum B, Reference for Causal Factors and Corrective Actions. Another good reference when completing this evaluation, in terms of its comments on human errors that may be made above the worker level, is Chapter 4 here. 

Most of the epidemiological data on the association of bladder tumors and the environment are centered on coffee consumption and artificial sweeteners. From case-control studies, there appears to be no causal relationship between coffee consumption and bladder tumor incidence (NAS, 1982 Omenn, 1982). Experimental animal models would predict that saccharin should promote tumors in the human. However, to this date there is no epidemiologic data that would indicate an increased tumor rate associated with populations consuming quantities of artificial sweeteners (Saccharin and Bladder Cancer, 1980 NAS, 1982). Perhaps the time frame is not correct or the dose is too low to demonstrate an effect in the human population. Further studies, especially on diabetic populations, are required to ascertain if artificial sweeteners can serve as promoters in human bladder cancer. 

Extrapolation of risk factors from rodent to humans is difficult because of differences in longevity, mobility, and the multistage development of cancer, but correlations can be made between dietary factors and tumor incidence. Based on experimental animal models and epidemiologic investigations, the combined evidence is strongly suggestive for a causal relationship between fat intake and occurrence of cancer for the breast, stomach, and colon. Data from animal studies suggest that increased total fat increases the... 

It is evident and convincing that several observers recognize the impact that management decisions have on what become causal factors for hazards-related incidents in the sociotechnical work environment— in the system in which work is done. Similarly, this author s review of over 1700 incident investigation reports results in the conclusion that decisions made at a board of directors and at senior management level must be given prominence in a causation model. Figure 13.1 depicts such a causation model. Comments in support of the model follow. 

---