ArcelorMittal’s lost-time injury rates have been declining steadily over the past 10 years, from 3.1 in 2007 to 0.78 in 2017. And while fatality rates have also improved over this period, over the past 5 years they have levelled off. How can we tackle this apparent contradiction and bring fatalities down to zero?

The data suggests that the traditional health and safety idea - that accidents which have resulted in minor injuries are a precursor to those which pose a risk of more serious injuries or fatalities - may not be as helpful as the classic health and safety pyramid would have us believe.

In other words, logging and analysing reactive, or lagging, indicators like lost-time injuries and restricted work injuries, may not lead to accurate assumptions about the likely causes of serious injury or fatality.

On the other hand, looking at leading indicators, such as the number of accidents, whatever their real consequences, with a specific potential for a severe injury or fatality (PSIF), means potentially dangerous situations can be dealt with before anyone is hurt and correlate directly with a reduction in actual serious injury and fatality.

Our data shows us this is in fact the case. Our fatality-free sites proactively detect and manage twice as many PSIFs compared to sites which have had one or more fatality; conversely, those sites which have suffered fatalities face 50% more actual lost-time and restricted-work injuries that had the potential for fatalities than their fatality-free peers. The challenge is to work with these sites to proactively detect more potentially unsafe situations, and so be in a position to manage them.

To make all our operations safer than ever, we are encouraging a focus on proactive PSIFs as a key performance indicator (KPI) - as well as other leading indicators of future safety performance. In practice, this means enabling people across the company to better record and benchmark potentially unsafe situations within our company-wide REX database, by:

• Providing safety training to reduce risk tolerance while fostering safety leadership
• Measuring the quality as well as the number of shop floor audits
• Assessing the value of pre-shift safety meetings conducted
• Evaluating the quality as well as the number of pre-job risk analyses undertaken onsite.

In the end, it’s not the improved PSIF data that prevents fatalities, but the habit of noticing the potential for severe injury to oneself or one’s co-workers, and flagging it up so that everyone can learn from the situation to avoid it happening again. That’s why, through training, we are helping employees develop the skills to recognize, log and examine what could go wrong. We believe this is what will make people across the company safe.