Recently, I’ve been thinking about risk assessments—and the many names we give them: JSA, JSEA, task analysis, SWMS, Take 5, STAR, and others.
We identify the hazards, assign a risk rating, apply controls, and then reassess the risk rating. But this raises an important question:
How can we confidently conclude that a risk rating has actually been altered, simply because controls have been identified? e.g. a risk that was initially considered likely is now unlikely.
This becomes particularly important when the controls rely heavily on administrative measures or PPE, where effectiveness can depend on human behaviour and consistent application.
Coincidentally, Safeguard Magazine recently published its periodic Safeguard Update, which included an article on “Better control decisions.” It referenced research by Yaqoob Raheemy, Fred Sherratt and Matthew R. Hallowell: What makes a safety control “enough” for high-energy hazards? Defining and testing acceptable alternative controls when a Direct Control is not feasible.
It’s an insightful read, and one statement regarding serious injuries and fatalities (SIF) particularly resonated with me:
“SIF events frequently arise from work contexts where controls are absent, ineffective, poorly specified, or implemented in a fragmented way, such that they fail to interrupt the pathways that lead to severe outcomes.”
This got me thinking about the effectiveness of risk assessments in our workplaces.
How many of our controls—particularly those addressing critical risks—actually interrupt the pathway to serious harm?
Perhaps we need to ask a different question when selecting controls:
What are we actually trying to achieve with this control?
An often-overlooked step in risk management is to risk assess the control itself. Is it strong enough? Is it reliable? What could cause it to fail? And, importantly, how do we know it is actually working as intended?
Identifying a control is not the same as controlling the risk.
Risk management is critical and widely practised. The bigger question we need to ask ourselves is: