What works in safety: The use and perceived effectiveness of 48 safety interventions

This surveyed members of the Dutch Society for Safety Science (NVVK) and their views on what safety interventions are used and what they believe are effective.

297 members responded to the survey (which had pre-defined responses and a free-text area), which consisted of four main parts relevant to this paper:

1.?????What does your company do regarding safety interventions?

2.?????What intervention has improved safety in your company most?

3.?????Why do you choose one intervention over another?

4.?????Describe an intervention that you would recommend to others and one you do not endorse?

Notably, they focused on general safety interventions and not interventions aimed at specific areas, like process safety or towards specific events, like spills or accidents.

Remember also that this study didn’t evaluate the actual efficacy of interventions, but rather peoples’ perceptions of what is/isn’t effective.

The interventions were rank ordered into three groups:

High (1 – 16), medium (17 – 32) and low (33 – 48). Next they categorised these into four groups based on their use and perceived effectiveness. The four categories were:

1.?????Interventions used relatively frequently and often perceived as effective (Hu-Hpe group)

2.?????Interventions used relatively frequently but rarely perceived as effective (Hu-Lpe)

3.?????Interventions used relatively rarely but often perceived as effective (Lu-Hpe)

4.?????Interventions used relatively infrequently and rarely perceived as effective (Lu-Lpe)

I can’t present all of the data, so you’ll need to hit up the paper.

Results

This highlights the interventions grouped into respective categories and frequency:

The figure below highlights the most common and least common top 10:

Data highlights that safety training; cataloguing and investigating incidents and near misses; systemic scheduled discussions of safety issues; and safety observations were common interventions.

There are also several interventions used relatively frequently that people perceive as not particularly effective (Hu-Lpe in the tables). This includes internal audits of the SMS, visual markings on the work floor, use of safety symbols on dangerous objects, machines etc., and providing site visitors with basic house rules.

Three interventions were considered effective but used infrequently (Lu-Hpe): use of a safety day, campaign focused on improving employee safety behaviour (posters and the like), management of change procedures.

Seven interventions are used relatively infrequently and also not seen to be particularly effective: using lessons from earlier projects in work preparation, accessible electronic system for managing safety action items, managing lists of safety critical elements, and human centred designs of computer systems to reduce human error.

When respondents gave main reasons for why they implemented a specific intervention – legal requirements were clearly an important stated consideration. This is supported by the first image above where “general risk inventory and assessment” is a key intervention category, intended in part to address legal requirements. This intervention group was also seen to be particularly effective by respondents.

Expectedly, different views were held on interventions. For instance, some held last-minute risk assessment in high esteem whereas others saw them more as tick and flick exercises.

Overall, it’s highlighted that “The safety practitioners in our survey have a clear preference for interventions which they think improve safety most, like employee training, management training and accident investigation” (p5).

In saying that, many commonly used interventions were also not seen to be particularly effective.

Link in comments.

Authors: van Kampen, J., Lammers, M., Steijn, W., Guldenmund, F., & Groeneweg, J. (2023). Safety Science, 162, 106072.