Safe As pod eps 10 & 11: How safety myths hold back progress and the flaws in our injury measures

This week on Safe As podcast (yes, a minor name tweak) we unpacked the role of safety myths from Besnard & Hollnagel in holding back progress and learning, and our misplaced faith in injury measures from Hallowell and colleagues.

Links to both eps below.

Safe As ep 10 explored the below safety myths – and proposed alternatives – image 1 below:

1. Human error is the largest single cause of incidents

2. Systems will be safe if people comply with the procedures they have been given

3. Safety can be improved by barriers and protection; increasing the layers of protection leads to higher safety

4. Root cause analysis can identify why mishaps happen in complex socio-technical systems

5. Accident investigation is the logical and rational identification of causes based on facts

6. Safety always has the highest priority and will never be compromised

Safe As ep 11 explored the statistical limitations in our use of TRIFR from Hallowell et al. – image 2, finding:

· TRIR is not associated with fatalities and “improvement in TRIR performance may not necessarily prevent fatalities”

· TRIR is almost entirely random and hence “recordable injuries do not occur in predictable patterns or regular intervals”, whereby “safety is a complex phenomenon that is impacted by many factors”

· TRIR cannot be represented as a single point estimate and is best expressed as a confidence interval and studied over time

· TRIR is not precise and should not be communicated to multiple decimal points unless there are hundreds of millions of worker-hours and suitably narrow confidence intervals

· If an organization is using TRIR for performance evaluations, they are likely rewarding nothing more than random variation, because statistically speaking, we cannot discern actual changes from chance variation

· TRIR is predictive only over very long periods of time, e.g. perhaps upwards of 100 months of data