Every serious safety event arrives with the same quiet feeling attached, the sense that somewhere in the building, someone already knew. Not the day it would happen, and not the patient. The weak spot. Someone knew that the shift-change handoff ran on one nurse’s memory, that the infusion alarm had been overridden so often it had stopped meaning anything, that the label printer down the hall produced stickers faster than anyone could match them to the right patient. The knowledge existed. It just never reached a room where someone had both the authority to act and the whole picture at once.
We tend to call these events unforeseeable, and the word does a lot of defensive work. There is a real difference between unpredictable and foreseeable, and most harm sits on the wrong side of it for comfort. Nobody could have predicted that this patient, on this Tuesday, would receive the wrong dose. Almost anyone who sat down and looked honestly at the process could have told you, weeks earlier, that the dose stood a good chance of going wrong in exactly that way. Predicting the instance is impossible. Seeing the vulnerability is ordinary work that we mostly decline to do.
Here is the part that is harder to say out loud. Most organisations are not under-invested in safety. They are heavily invested, in one direction only. They report incidents, they convene reviews, they write corrective actions, and every piece of that work happens after someone has already been hurt. It is necessary work, and I would never argue against it. But an organisation whose entire safety effort points backward has made a choice, whether it realises it or not, to keep meeting the same foreseeable failures with the same expression of surprise. Reacting well can feel like diligence while quietly functioning as a way of never having to look forward.
Looking forward has a name and a method. Failure Mode and Effects Analysis is, underneath the acronym and the worksheets, a structured way of asking four plain questions about a process while it is still working. Where could this go wrong? How badly? How often? And if it did go wrong, would anyone catch it before it reached the patient? A group of people who actually do the work sits down, walks the process as it really happens rather than as the policy describes it, and answers those questions before the incident report is needed. The output is not a binder. It is a short list of the weak spots that matter most, each with a name and a date attached to it.
The honest objection is time, and it deserves a straight answer rather than a slogan. A full analysis costs real hours, and those hours have to come from somewhere. Pretending otherwise insults the people you are asking to spend them. But the arithmetic is not close. A single serious event consumes more staff time in its aftermath, across review meetings and disclosure conversations and corrective plans, than the analysis that would have surfaced the hazard in the first place. The proactive hour is not free. It is simply the cheapest one available.
None of this requires a quality department, a certification, or a software licence. It requires a process worth worrying about, the right handful of people in one room, and the willingness to ask the uncomfortable question before circumstances ask it for you. The surprise on the day of the event is the one part of this whole story that was always optional.
This is one of the ideas at the centre of my book, Before Harm Happens: Medical FMEA in Practice, which turns that four-question habit into something any healthcare team can run, from a thirty-minute huddle to a full analysis. If the argument here rings true, the book is where the how lives.
This article accompanies the book Before Harm Happens.