If We Stop Blaming People, Don't We Lose Accountability?

At some point in almost every conversation about becoming a "blame-free" hospital, someone asks the question everyone else is already thinking:

"So what happens when someone actually screws up? Do we just say systems and move on?"

It is a fair question.

A safe culture cannot mean that every mistake is excused or every behavior is treated the same. Accountability still matters. If every event is explained away with a sympathetic nod and a list of contributing factors, teams will eventually lose confidence that meaningful problems will actually be addressed.

The person asking this question is not standing in the way of safety culture. Often, they are trying to protect the standards that matter.

The problem is the assumption underneath the question: that accountability requires blame.

It does not.

Blame asks who. Accountability asks what.

Blame is retrospective and personal. It wants a name. Its emotional function is to restore a sense of order: something bad happened, someone must have caused it, once we identify them the world makes sense again. Blame feels like justice and often produces nothing.

Accountability is prospective and structural. It asks what happened, what conditions allowed it, what we are going to do about it, and who owns that change. It is not softer than blame. In many ways it is far more demanding, because it doesn't let anyone off with an apology. It requires that something actually change, and that someone answer for whether it did.

You can have a fiercely accountable hospital with almost no blame in it. What you cannot have is a hospital that blames constantly and learns anything.

The line most hospitals never draw

A just culture draws a distinction that most veterinary teams have never made explicit. It sorts behavior, not outcomes, into three categories, and each one gets a different response.

Human error. The slip, the lapse, the honest mistake. The nurse who reached for the vial she has reached for two hundred times and this time it was the wrong one. The clinician who transposed two digits at the end of a fourteen-hour day. This is what competent, well-trained people produce under normal conditions, and no amount of exhortation eliminates it.

The response is to console, and to fix the system. You support the person, and you change the conditions that made the error easy. Punishing human error is not just cruel; it is functionally useless, because the next person will face the identical setup.

At-risk behavior. The drift. The shortcut that has become normal because it saves four minutes and has never yet caused harm. Skipping the second check when the room is slammed. Pre-drawing syringes and labeling them later. Nobody chose to be unsafe. The risk simply stopped being visible, usually because everyone does it and nothing bad has happened.

The response is to coach, and, critically, to ask why the shortcut exists. At-risk behavior is almost always a rational adaptation to a badly designed process. If the safe way takes twice as long as the unsafe way, you have a design problem wearing a behavior costume.

Reckless behavior. The conscious disregard of a known, substantial, unjustifiable risk. Not a slip, not a drift, but a choice made with awareness of the danger. Operating impaired. Falsifying a record. Deliberately bypassing a control the person knows is there for a specific reason.

The response is discipline. This is rare. In most hospitals it is genuinely rare. But it exists, and a just culture says so out loud, because a framework with no consequences at any point is not credible to the people you most need to convince.

The unspoken rule that fills the vacuum

Here is what happens when a hospital never draws that line.

The vacuum does not stay empty. It fills with a single unspoken rule, absorbed by everyone within their first month:

If the outcome was bad, someone is in trouble.

Nobody writes this down. Nobody would defend it if challenged. But it governs behavior completely, and it produces exactly the results you would predict. People hide the near miss, because a near miss with a good outcome is a risk with no upside. They soften the report, not falsifying it, just choosing the gentler verb. They wait to see if anyone noticed before deciding what to say. They manage their own exposure, rationally and continuously, because nothing in the environment has told them what the actual standard is.

You do not get accountability that way. You get silence, and silence is indistinguishable from safety right up until the moment it isn't.

The published veterinary work points at the mechanism underneath this. Recent research on error disclosure has found that professional identity is deeply entangled with error, in the belief that a good clinician is one to whom things do not go wrong. When that is the operating assumption, disclosure is not merely embarrassing; it is a threat to the person's sense of who they are. Work on humble leadership and error disclosure in Australian practice suggests the reverse also holds: when senior clinicians model fallibility openly, the cost of disclosure for everyone below them drops sharply.

The part worth sitting with: outcome is not the measure

This is where a just culture asks something genuinely difficult of leaders, and where most implementations quietly fail.

The same action can deserve a conversation or a consequence depending entirely on the choices around it, not the size of the harm.

Consider two technicians.

The first follows every protocol correctly. Right drug, right dose, right patient, second check completed and documented. The patient has an idiosyncratic reaction and dies. It is a catastrophe, and the technician did nothing wrong.

The second skips the safety checks entirely, deciding the patient is stable, the room is busy, and the check is a formality today. Nothing happens. The patient walks out fine that afternoon.

Every instinct in a hospital pulls toward investigating the first case and never hearing about the second. The first has a body. The second has nothing to point at.

But the behavior that needs your attention is in the second case, and the person who needs your support is in the first. Judge the behavior, not the luck of the outcome.

This is hard because it runs against how humans actually assign fault. Outcome bias is well documented and it is not a failure of character. The severity of a result genuinely changes how we perceive the decision that preceded it, and it does so automatically. Knowing about the bias does not switch it off. The only reliable defense is a process that evaluates the choices before the outcome is allowed into the room. Some teams do this literally: describe the decision sequence first, discuss it, and only then reveal what happened.

What changes when you say it out loud

Draw that line explicitly, in writing, in orientation, in how you actually respond, and something shifts that is difficult to achieve any other way.

People stop managing their exposure and start telling you the truth.

Not because they suddenly became braver, but because the calculation changed. When the standard is knowable in advance, disclosure stops being a gamble. The technician who skipped the check knows what that conversation will be. The technician who followed every protocol and lost a patient knows they will be supported rather than investigated. Both of those certainties are worth more to your reporting culture than any amount of reassurance.

And the senior clinician who asked the hard question at the meeting gets a real answer: that this framework is more demanding than what you had before, not less. It just aims the demand at the right target.

What to do this week

Write the three categories down. One page. Error, at-risk, reckless, with a veterinary example of each drawn from your own hospital's history. If you cannot produce an example, that is informative.

Say the reckless part out loud. Teams do not trust a framework that appears to have no floor. Naming the small number of things that will always meet consequences is what makes the rest of it believable.

Separate the two questions in every review. First: what were the conditions? Second, and only second: what were the choices? Running them together is how outcome bias gets in.

Check whether your team could pass a quiz on this. Ask three people at different levels to describe the difference between an honest mistake and a reckless one. If you get three different answers, you have not drawn the line yet. You have only thought about drawing it.

A question for your team

Could everyone in your hospital describe the difference between an honest mistake and a reckless one, and do they trust you to tell them apart?