Watch what happens at the end of most risk analyses. The hard work is done, the hazards are named and ranked, and then, in the last forty-five minutes, the team disposes of them. The disposal, with dispiriting regularity, reads the same way. Re-educate the staff. Revise the policy. Remind everyone at the huddle. Add it to orientation. Everyone leaves satisfied, and almost nothing changes.
None of those actions is worthless. What they share is a fatal property: they try to fix the process by fixing the people, and the people were rarely the problem. The nurse using the workaround already knows the policy. The workaround exists because the policy collides with reality, and a memo does not move reality.
Human factors work gives us a way to rank interventions honestly, by how much they lean on a tired human remembering to do the right thing every single time. At the strong end are forcing functions, where the wrong action becomes physically or logically impossible: the connector that only fits one outlet, the oral syringe that will not attach to an intravenous line, the order that cannot be filed without the weight it needs. Next, automation, where the system performs or checks the step without depending on vigilance. Then simplification and standardisation, fewer steps and fewer variants and fewer decisions to get wrong. Then, weaker, the reminders and checklists and double checks, genuinely valuable and genuinely fragile, because each one spends a little of a finite supply of attention. And at the bottom, necessary as a foundation and feeble as a corrective action, the rules, policies, and education that assume knowledge was the missing ingredient when it almost never was.
The point of the ranking is not to forbid the bottom of it. Sometimes a reminder is all that is affordable this quarter, and that is an honest answer. The point is a discipline: for every hazard, the team has to say out loud why it is not acting higher up the list. Cost, vendor limits, timeline, fine, write it down, take the weaker action now, and file the stronger one where the next budget cycle will trip over it. What the discipline prevents is the lazy slide to education because it is cheap, fast, and needs nobody’s permission.
There is a blunt corollary worth keeping in view. The strength of the fix should match the severity of the hazard. A catastrophic, poorly detectable failure answered with a laminated reminder is not an action plan. It is a document that will read very badly later, to exactly the audience you least want reading it.
So when the action phase arrives, slow it down for a moment and ask the awkward question of each proposed fix. Are we changing the process, or are we asking people to be more careful inside a process we have left exactly as dangerous as we found it? The honest answer is uncomfortable more often than anyone would like, which is precisely why it is worth asking.
This article accompanies the book Before Harm Happens.