Separating the act from the fiction
October 1st 2007

In Industrial Accident Investigation what are 'unsafe acts' and how can they be remediated? Paul Difford of the Institute of Industrial Accident Investigation took this question to the audience of Health+Safety 07 as part of his seminar on Industrial Accident Investigation:

Since a leading study in the 1920's, 'unsafe acts' continue to be identified as the immediate cause of many industrial accidents. But what are they? Organisations often provide lists from which their accident investigators can 'select' the type of unsafe act that they have found; many use terms like 'failed to wear PPE', 'using equipment inappropriately', 'failed to follow procedures' or 'failed to secure'.

Typically, the remedy will be a reminder of what should have been done and/or a recommendation for re-training.

But, such 'remedies', especially when re-training simply involves re-attending a previously attended course, can be a dangerous waste of resource. As essentially broad descriptions of cause, such lists of unsafe acts can inadvertently appear to be classifying behaviour; this is because wording like 'failed to wear PPE' can suggest that those involved intended or chose to behave as they did.

But, some of the behaviour that takes place on our sites requires little by way of conscious control to maintain it.

For instance, steel fixing is a skill which most of us find difficult unless we apply a great deal of constant mental effort; even then, the result is usually a waste of time and wire. Yet, the skilled fixer, balancing on an uneven mat, can look away from the work, talk and joke with colleagues and maintain a seemingly effortless pace.

Skill-based behaviour This side of their work involves skill-based behaviour and is performed almost automatically due to its highly practiced and well learned nature. Once underway, it requires little conscious effort. However, it does require checks on progress by the individual and it is during such checks that any error is usually detected and corrected.

In the main, the detection of an error (e.g. continuing with a type of tie when a change was required) by a fixer is likely to result in annoyance rather than something sinister.

But, the error, the skill-based error, will not have arisen from any intention or conscious decision to install 200 wrong ties that now need replacing; therefore, a reminder on what should have been done is largely pointless.

More often than not, a skill-based error can only be detected and corrected in time by the individual who has made it. Unfortunately, during certain skill-based tasks, the delay in making a check on progress (i.e. detecting the error) can expose an individual (and the organisation) to the chance of a fatal skill-based error; but, as already stated, it is not an unsafe act in the sense of the examples provided above because it involves no conscious decision to commit it. Consequently, the remedial intervention strategy that has not taken account of this will be ineffective.

Example 1 As an example an organisation requested an assessment of the systems and training in place for its slinger/banksmen following the identification of a trend during accident data analysis. The accidents concerned had been categorised as both 'failed to secure' and 'failed to follow procedures'. In all cases, remedial intervention consisted of either a reminder of 'procedure' or a requirement for re-training.

Incidentally, some areas of the business had blamed the 'failure to secure' on the 'failure to follow procedures' and concluded that this was evidence of poor culture and contractor management. However, evidence of skill-based error in some of these reports was not pursued owing to the limitations of the definitions/classification ('failed to secure' etc) and the investigators inability to highlight it and analyse it.

Example 2 In one report, a highly experienced slinger/banksman had successfully attached one leg of a lifting chain to a stillage of scaffold tubes. While attaching the other leg his mobile phone rang and he answered it. Stopping the job, he strolled about aimlessly whilst talking and by the end of the call was on the other side of the stillage. Responding to a shout from the crane driver, he signalled for the slack to be taken. All looked well (because the unattached leg had snagged the stillage) and so he signalled for the load to be lifted. Twenty-five feet into the air, the unattached leg let go and the tubes fell from the stillage.

It was found that this slinger/banksman had 'slipped through the system' and was not in possession of a current ticket. The unsafe acts were listed as 'failed to secure' and 'failed to follow procedures' and were remedied by sending him for training.

These example leave us with two questions. What was the cause of that accident? And will training necessarily prevent, or reduce the likelihood, of the next one?

Health+Safety 07 - Seminars to download

To download the full presentation Industrial Accident Investigation: The Problems with Casual Analysis and the Classification of 'unsafe acts' or any other presentation from Health+Safety 07 visit www.healthandsafetyevents.co.uk

The next installments of this regional show which features a free conference and seminar programme are:

26-27 February 2008, Sandown Park, Esher 8-9 October 2008, The Reebok Stadium, Tel: 01932 568344 Bolton

More articles from Principal People: