Somewhere tonight, in an operating room, an ICU or a step-down unit, someone is going to look at a patient and feel that something is not right. Maybe they cannot explain it yet. Maybe the numbers have not caught up with what they are seeing. But they notice.
And then there is that half second. The moment when they decide whether to say something out loud or wait. Whether the concern is important enough. Whether they will be dismissed, embarrassed, or made to feel that they should have known better. Whether speaking up is worth what might come next.
That half second is what this article is about. In healthcare, we measure almost everything. Door-to-balloon time. Time to antibiotics. Length of stay. Readmissions. First-case on-time starts. We have become remarkably good at measuring what happens.
What we are not nearly as good at measuring is what never happens.
We do not measure the question someone decided not to ask. The concern someone convinced themselves could wait. The nurse who noticed something but did not feel comfortable calling the surgeon again. The resident who hesitated because they did not want to look inexperienced.
And we certainly do not measure the sentence that might have changed what happened next, but was never said.
That is where psychological safety stops being a conversation about culture and becomes a conversation about patient safety.
What the data says
A 2019 study in JAMA Surgery looked at 13,653 patients operated on by 202 surgeons across two academic medical centers. The researchers linked each operation to something most of us have never treated as clinical data: the reports coworkers had filed about that surgeon’s behavior in the preceding 36 months.
Patients whose surgeon had four or more coworker reports had a complication rate of 14.1%. Patients whose surgeon had none had 10.7%. Surgical complications rose. Medical complications rose. Surgical site infections rose. Sepsis rose. Read that last one again. Sepsis rose.
The authors are careful, and so should we be. This is an observational study. It does not establish causation, and they say so plainly. But it belongs to a growing body of work, including randomized simulation studies, pointing in the same uncomfortable direction: how a team is spoken to changes how a team performs, and the patient is downstream of both.
What this is not
I do not believe people come to work intending to be difficult. Healthcare is hard in ways that are difficult to explain unless you have lived it. We take care of people on some of the worst days of their lives. We make consequential decisions under enormous pressure, sometimes after being awake far too long, and often while working inside systems that make doing the right thing harder than it should be.
So when someone becomes frustrated or loses their composure, I think we have a responsibility to understand what happened before we decide what it means. Was the team short staffed? Did the equipment fail at exactly the wrong moment? Was this the fifth interruption in nine minutes? Was someone trying to care for a very sick patient while fighting a system that was not helping them do it?
Psychological safety has to work in both directions. The person whose behavior is being questioned deserves the same opportunity to be heard and understood as the person who experienced it. But understanding behavior does not mean accepting it.
Medicine has a long history of making exceptions for people because they are technically exceptional. We tolerated the surgeon who raised their voice, humiliated a trainee, or made everyone in the room afraid to speak because they were brilliant, productive, or the person everyone wanted when the case became difficult.
I think we know better now.
Technical excellence cannot excuse behavior that makes it harder for someone else to speak when a patient may be in danger. This is not about whether someone is nice or difficult. It is not even primarily about professionalism.
It is about whether the people around us feel safe enough to tell us when we might be wrong. And that is patient safety.
The real question
The question was never whether someone yelled. The question is what happens the next time that same person sees something wrong. The study describes exactly this scenario: a clinician reports that a patient is hypotensive, has started pressors, notifies the surgeon, and is yelled at for five minutes. The authors make the point without drama. That clinician may hesitate the next time a patient deteriorates.
Hesitation is not insubordination. It is arithmetic. People run a fast calculation about cost, and humiliation is expensive.
You can be a superb technical surgeon. None of us operates alone. The safest person in that room is surrounded by people willing to tell them what they cannot see from where they are standing. If the people around you are afraid of you, they will give you less information at exactly the moment you need more of it, and they will give it to you later than they should have. The authors frame high reliability as depending on “communication, mutual respect, and continuous situational awareness.” Every one of those is a team property. Not one of them can be achieved alone.
What leaders owe
What do we do with this? First, we need to teach psychological safety the same way we teach any other skill that affects patient care. We train clinicians in intubation, escalation pathways, and closed-loop communication during a code because we know those skills change outcomes. Psychological safety changes outcomes too. It should not be something we talk about once a year at a retreat, and it certainly should not be introduced only after someone has already had a behavioral event. It needs to be part of how we teach people to work together from the beginning.
We also need to start treating behavioral reports for what they are: safety data. Too often, we dismiss them as personality conflicts, interpersonal problems, or something for HR to manage. But when multiple coworkers are raising concerns about the same person, that pattern is telling us something. If that signal appears years before a complication, we cannot afford to ignore it simply because it is uncomfortable to address.
And perhaps most importantly, we need to stop confusing avoidance with kindness. When someone repeatedly undermines a team and nobody has the courage to tell them, we are not protecting that person. We are actually denying them the opportunity to change. At the same time, we are asking everyone around them to absorb the consequences and potentially transferring that risk to our patients. The encouraging part is that people can change. The literature shows that peer-delivered, tiered interventions can reduce subsequent reports, even among people with persistent patterns. Most people, when approached clearly, respectfully and early, will course correct.
There is one finding I keep coming back to. Female surgeons were substantially underrepresented among those generating coworker reports. The study does not tell us why, and I do not want to make the data say something it does not. But I do think it should make us curious. As leaders, we should be willing to ask not only whose behavior we correct, but whose behavior we have historically learned to tolerate, who receives the benefit of that tolerance, and ultimately, who we choose to advance.
The part that is not about data
I keep thinking about the person who does not speak up. And I do not mean the surgeon. We spend a lot of time talking about the surgeon. I mean the scrub tech who noticed that the count did not seem right but convinced herself it was probably nothing. The resident who saw the blood pressure drifting in the wrong direction but decided to wait for the next set of vitals before saying something. The nurse who finished her shift, walked to the parking garage, sat in her car and needed a few minutes before she could turn the key because she almost said something and did not.
That person stays with me. Because if something goes wrong, she will replay that moment. She may carry it for years. And the uncomfortable truth is that she was not imagining the cost of speaking up. Somewhere along the way, we taught her that there could be a cost. Maybe nobody ever said it directly, but she learned it from the way we responded to questions, from the tone in the room, from watching what happened to someone else who challenged a decision.
That is the part we cannot capture in a spreadsheet.
The patient lying on the table knows none of this. They do not know who has the endowed chair, who has published the most papers, who has the biggest reputation, or who has done the most cases. None of that matters to them in that moment. They are asleep, trusting a room full of people they may never remember meeting to communicate with one another when it matters.
Sometimes keeping that patient safe comes down to one person being willing to say one sentence: “Something is not right.”
Our responsibility as leaders is to create a culture where saying that sentence does not require fear.
Dr. Ferrada is chair of the department of surgery and medical director of perioperative services at Inova Fairfax Medical Campus, professor of medical education at the University of Virginia, and division and system chief of trauma and acute care surgery at Inova Health System.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
