In the mid-1990s, a young Harvard researcher named Amy Edmondson set out to show that better hospital teams made fewer medication errors. Her data showed the opposite. The units with the best teamwork and leadership were reporting more errors, not fewer.
She had not found that good teams were careless. She had found that good teams were willing to talk about their mistakes. The paper, published in the Journal of Applied Behavioral Science in 1996, became one of the starting points for the idea now known as psychological safety.
That origin story matters, because psychological safety has since drifted into something softer. In many hospitals it is now discussed as if it meant a pleasant atmosphere, supportive language and fewer raised voices. Those are fine things. They are not the point, and in the operating room the confusion can be dangerous.
What the concept actually describes
Psychological safety is a shared belief within a team that it is safe to take interpersonal risks: to ask a question, admit an error, point out a problem or disagree with someone senior. It is about whether people speak up. It is not about whether they feel comfortable.
In fact, teams with high psychological safety often feel less comfortable, because more problems are being surfaced and argued about. The quiet, harmonious theatre where nobody ever questions the surgeon may be the least safe room in the building.
A scenario worth rehearsing
Picture the end of a long abdominal case. The scrub nurse’s count is one swab short. The surgeon is closing, the anaesthetist is thinking about the next patient, and the list is running late. The nurse is fairly sure the count is off, but not certain; swabs have been miscounted before.
In one theatre, she says so, the surgeon pauses closure, the team searches, and the swab turns up on the floor under a drape. Four minutes lost. In another, she hesitates, recounts in her head, and decides not to hold up a consultant who snapped at her last week. Most of the time nothing bad follows, because most of the time the count really was wrong. Occasionally it does, and a patient returns weeks later with a retained item.
What decided the outcome was not the nurse’s competence, or the existence of a count policy. Both theatres had the policy. It was her prediction of what would happen when she spoke. That prediction was built from dozens of earlier moments, most of them far less dramatic, in which she saw how concerns were received. A sigh, a sarcastic reply, a consultant who later told a colleague that she was “always fussing”: each of these is small, and each is remembered. Psychological safety is the accumulated residue of those moments, which is why it can be destroyed in a week and takes months to rebuild.
The operating room is a hard case
Surgery combines everything that makes speaking up difficult. The hierarchy is steep and visible. The work is time-pressured. The person most likely to be making a consequential decision is also the person with the most power in the room. And junior staff, scrub nurses and trainees know that their future references, rotations and working lives depend partly on how that person regards them.
The WHO Surgical Safety Checklist was designed partly to flatten this for a few critical moments. In the landmark study published in the New England Journal of Medicine in 2009, led by Alex Haynes, introducing the checklist across eight hospitals in eight countries was associated with a fall in in-hospital deaths from 1.5% to 0.8%, and in complications from 11.0% to 7.0%. One reason given for the effect is that the checklist creates structured moments where everyone, including the most junior person in the room, is expected to speak.
But the results have been harder to reproduce in routine practice. A 2014 study in NEJM led by David Urbach, covering 101 hospitals in Ontario, found no significant reduction in deaths or complications after mandatory checklist adoption. One plausible interpretation is that a checklist can be performed without being believed. If the circulating nurse reads the items aloud while the surgeon is already scrubbing and nobody waits for answers, the tool becomes ritual.
There is a second reading of the Ontario result that deserves mention, because it cuts in a more hopeful direction. The Ontario hospitals were already performing well. Adjusted mortality in the study was 0.71% before the checklist and 0.65% after, against 1.5% at baseline in the original eight-country study. Some of the hospitals in the original study had far more room to improve. A tool that makes a large difference where basic communication is weak may have little measurable effect where it is already reasonably strong, and the Ontario study, which compared three-month windows either side of adoption, may simply have been too short and too crude to detect the change in behaviour.
Both readings lead to the same conclusion. The checklist is not the active ingredient. The conversation it is meant to force is. Where that conversation already happens, the paper adds little. Where it does not happen, the paper alone will not create it.
That is the trap. Structures that invite speaking up do not work unless people believe speaking up will be welcomed.
Why new technology exposes the gap
Edmondson’s later work with Richard Bohmer and Gary Pisano, published in Administrative Science Quarterly in 2001, studied sixteen hospitals adopting minimally invasive cardiac surgery. It was a technique that disrupted established routines and made the surgeon more dependent on the rest of the team. Only some of the hospitals implemented it successfully. The successful teams were deliberately selected, prepared together and reflected together after cases. In those teams, surgeons actively invited input from nurses, perfusionists and anaesthetists.
This matters more now, not less. Robotic surgery, new imaging and AI-assisted tools all change the flow of information in the operating room. The surgeon at a console may see less of the room than before. The team’s ability to raise concerns quickly becomes part of the technology’s safety case.
The objection from the surgeon’s side
Some surgeons hear all this and worry, reasonably, about the opposite failure. An operating room in which everyone voices every doubt could become slow and hesitant, and in a crisis, such as a major bleed, the team needs one clear voice, not a seminar. Decisiveness is not a character flaw in surgery; it is sometimes the thing that keeps the patient alive.
That objection partly holds. Psychological safety does not mean every decision is shared or every concern must be debated. It means concerns reach the person making the decision, quickly and clearly, so that they decide with the information. The surgeon still decides.
The practical answer is structured escalation language, which lets a concern be raised quickly and understood instantly. The TeamSTEPPS programme developed with the US Agency for Healthcare Research and Quality, for example, teaches the “CUS” words: I am concerned, I am uncomfortable, this is a safety issue. The escalating phrases tell the listener how serious the speaker thinks the problem is, which is exactly what a decisive leader needs to know to decide whether to stop. Shared vocabulary makes speaking up faster, not slower.
How would you know?
Many hospitals now run staff surveys that include questions about whether it is safe to raise concerns. Those surveys are useful, but they measure what people say about speaking up, not whether they do it. And the teams with the biggest problem are sometimes the ones least likely to answer honestly.
Behavioural signals are harder to fake. How often do theatre teams actually pause a case after a concern is raised? How many near-misses are reported, and from which staff groups? Do debriefs happen, and do they ever record a problem? A theatre suite where nurses and trainees never report anything is not necessarily safe. Edmondson’s original finding still applies: silence can be a symptom.
The uncomfortable incentive
If psychological safety is so valuable, why is it so fragile?
Because surgeons are selected and rewarded for decisiveness, confidence and technical mastery. Those qualities save lives. They also make it harder to hear “I think we should stop.” And the cost of listening is immediate and personal: a pause, a moment of uncertainty in front of the team, a possible loss of face. The benefit is statistical and delayed, a complication that did not happen.
Nurses and trainees face their own calculation. Speaking up and being wrong can be embarrassing. Speaking up and being right, in some teams, can be worse.
So the burden has to sit with the most senior person in the room. It is unreasonable to ask the most vulnerable member of a team to take the biggest interpersonal risk.
What senior surgeons can do tomorrow
A few habits make a real difference to whether people speak:
- Say it explicitly at the start. “If you see anything that worries you, tell me. I would rather stop than miss it.” It sounds trivial, but it changes what the most junior person believes is allowed.
- Respond well the first time. The first time someone raises a concern, the reaction teaches everyone in the room whether it is worth doing again. Thank them even when they are wrong.
- Admit your own uncertainty. Leaders who say “I am not sure about this” give others permission to be unsure too.
- Debrief briefly after every case. Two minutes on what went well and what did not builds the habit of talking about problems before they become incidents.
Beyond the individual surgeon
Individual habits matter most, but organisations shape them. Hospital leaders can make it clear that stopping a case for a safety concern will never be punished, and back that up the first time a list overruns because someone did. They can include team behaviour in how surgeons are appraised, not only outcomes and volume. Surgical training programmes can teach speaking up and responding to it as core technical skills, practised in simulation where the stakes are low. And the leaders of surgical departments can pay attention to who is repeatedly described as difficult to raise concerns with, because those patterns are usually known long before they appear in incident reports.
The next few years will test this more, not less. As robotic systems spread and teams reorganise around consoles, screens and new roles, the old informal routes by which concerns reached the surgeon will change. Hospitals that treat team communication as part of the implementation of new technology, as the successful cardiac teams in Edmondson’s research did, will have a safety advantage over those that treat it as an afterthought. Buying the robot is the easy part. Rebuilding the team’s habits of talking around it is the work.
The takeaway
Psychological safety in the operating room is not a culture initiative or a poster in the staff room. It is a practical safety control, as real as instrument counts, and like any control it can be present on paper and absent in practice. The test is simple: when the most junior person in theatre sees something wrong, do they say so, and what happens next?
What is the one thing a surgeon has done that made it easier for you to speak up in theatre? Tell us in the comments.
