Use Observations of Unsafe Behaviors as a Leading Indicator

To know if you’re doing well, don’t look where you are. Look instead where you’re going. Even if your personal health indicators are good, it doesn’t mean that you’re doing well if you started to eat poorly or sleep less. The same applies to occupational safety. Lagging indicators (injuries, days away from work, etc.) give an organisation a glimpse of the past. They don’t reveal whether safety is improving.

Leading indicators provide a better idea of safety performance. They measure proactive, preventative, and predictive actions, which allows organizations to continuously improve safety. Leading indicators signal where things are headed. Examples include near miss reporting, participation in safety training, and observations.

The number of reported observations of unsafe behaviors can be used as a leading indicator because observations may reveal workplace hazards that can lead to risks of incidents. These risks can then be mitigated to improve safety.

Lagging Indicators Are Still Useful

The importance of leading indicators over lagging ones is accepted by many. But be careful not to dismiss lagging indicators. First, some regulations require organizations to report on incident rates and other lagging metrics. Second, lagging indicators help to measure outcomes to see if the activities measured by leading indicators are effective.

The comparison of the number of observations of unsafe behaviors and incident rates is a good example of the correlation between leading and lagging indicators. By comparing the number of observations and incident rates over time, you can have a good idea if observations are providing valuable insights into hazards. Note that there may be a gap between the evolution of the indicators. For example, if the number of observations has been high since January, incident rates may start to decline only as of April, since it may take time to fully implement control measures.

There may be different conclusions from the correlation between the number of observations and incident rates. Here are some examples:

• Observations are high, incidents are low: Observations may be successfully identifying hazards, which are being successfully addressed to reduce risks of incidents.
• Observations are low, incidents are high: Low number of observations may mean that hazards are not being identified and addressed, which is leading to high incident rates.
• Observations are high, incidents are high: Incidents are high because observations are of poor quality and not helping to identify hazards.
• Observations are low, incidents are low: Low number of observations may mean that safety performance has improved so much that there are fewer new hazards to identify. Low incident rates also reflect improved safety performance.

In addition, note that correlation may not imply causation, for example:

• Observations are high, incidents are low: Maybe most hazards were already known because of regular hazard assessments conducted independently, and have been addressed; while many observations may not be valuable and aren’t revealing new hazards.
• Observations are high, incidents are high: Observations may be successfully identifying hazards, but control measures are not successfully reducing risks of incidents. The problem is with risk mitigation, not observations.

Also, simply aiming for a higher number of observations may not produce the outcome that companies are looking for. They should focus on encouraging and rewarding participation and good-quality observations. This will result in greater buy-in from workers and help strengthen the safety culture.