Why Did Surgery at Kyoto University Hospital Not Stop?

On August 7, 2026, Kyoto University Hospital announced that a medical error had occurred during brain-tumor surgery: tissue from a region without the tumor had been removed.

The most disturbing feature of this case is not that the surgical team lacked information signaling that something was wrong.

It is that two test results indicated that the tissue was not a tumor, yet the information did not have the power to stop the operation.

The warning arrived.

But it did not function as a warning.

Unless we examine that difference, we risk confining the cause to the personality of one individual—to the “complacency” or “assumption” of a veteran surgeon—while leaving the same structure intact in another operating room.

The case asks more than why a physician made a mistake.

Once we accept that human judgment can be wrong, at what point, by whom and on what evidence can that judgment be stopped?

This is a question of medical technique, but it is also a philosophical question about knowledge, authority and responsibility.

This article is based on information made public by August 9, 2026. Kyoto University Hospital has established an accident investigation committee that includes outside experts, and the cause has not yet been determined. The analysis below does not claim to know the surgeon’s state of mind or undisclosed conditions inside the operating room. It considers what kind of system could have stopped the accident, based on the sequence of events that has been made public.

What happened during the brain-tumor operation?

According to the hospital’s announcement and news reports, the patient was a woman in her fifties. She had experienced dizziness and unsteadiness before surgery but had been living an ordinary daily life.

The operation was intended to remove an approximately three-centimeter benign tumor in the cerebellopontine angle, between the cerebellum and the brainstem. A meningioma was suspected.

The surgeon was in his forties, had more than twenty years of medical experience and had performed approximately one hundred similar operations.

During the operation, however, pathological examination of tissue believed to be part of the tumor found no clear evidence of tumor. The surgeon reportedly considered that the sample might have come from the edge of the mass and therefore might not contain tumor tissue, and continued the resection.

A second examination also indicated that the tissue was not a tumor, but the operation continued.

After an operation lasting approximately ten hours, the patient did not regain consciousness or spontaneous breathing. An MRI then showed that the intended tumor remained untouched, while the right cerebellar tonsil and the brainstem—areas involved in breathing and movement—had been severely damaged.

The patient remained in a serious condition on a ventilator, unable to move her limbs, and her prospect of recovery was reported to be unknown. Kyoto University Hospital’s announcement, Kansai TV and FNN’s detailed report, TV Asahi’s report on the hospital’s news conference

It is tempting to ask why two test results were “ignored.”

But the word ignored does not fully explain the danger in the judgment.

The crucial point may not be that the results went unseen. It may be that the results were seen and then given a meaning that preserved the original judgment.

“This is not a tumor” became “the sample must have been taken from the edge, so the tumor was missed.”

Evidence capable of refuting the hypothesis was transformed into material for protecting it.

Popper: a hypothesis that cannot stop action has not truly been tested

The philosopher Karl Popper located the strength of science not in its possession of absolute truth but in its willingness to expose itself to the possibility of being wrong.

A scientific hypothesis needs a condition of refutation: “If this result occurs, I will revise my belief.” If every possible result can be absorbed by an explanation added afterward, the hypothesis may appear to have been tested while in fact being protected from every test. Internet Encyclopedia of Philosophy, “Karl Popper: Philosophy of Science”

A judgment made during surgery can also be understood as a hypothesis joined to an action.

“The tissue in front of me is the tumor.”

“It is safe to continue resecting in this direction.”

On the basis of these hypotheses, a surgeon performs acts that cannot be undone.

Intraoperative pathological examination naturally has limits. A single negative result does not always rule out a tumor, and a specimen may miss tumor tissue depending on where it is taken.

The point, therefore, is not to declare from outside the operating room that every negative result must automatically end an operation.

The deeper question is whether the team had determined in advance what event would trigger a pause and a re-examination of both the anatomical location and the working hypothesis.

The first result might be an accident of sampling.

What, then, should happen after the second?

When visual assessment and pathology conflict, who performs the reassessment?

Does the same decision-maker evaluate the same hypothesis again, or is an independent judgment introduced?

Before the operation can continue, which contradiction must be resolved, and by what evidence?

If these conditions are undefined, a test can produce information without changing action.

A test with no stopping condition is not a safety device. It becomes merely another instrument displayed beside the operation.

Why did “one hundred similar operations” not guarantee safety?

The surgeon in this case reportedly had experience with approximately one hundred similar operations.

That number rules out the simple explanation that inexperience alone caused the accident. At the same time, it challenges the common assumption that more experience automatically produces safety.

In Black Box Thinking, Matthew Syed contrasts organizations that investigate failure and convert it into learning with organizations that protect prior beliefs by ignoring or reinterpreting unwelcome evidence. The latter become trapped in a closed loop: experience accumulates without genuinely testing the conclusion drawn from it. Discover 21’s Japanese edition of Matthew Syed’s book

Experience should be a store of material for revising judgment.

But repeated success can make experience perform another function.

“I have always been able to identify this anatomy.”

“A pathology result was negative before, and the tissue still proved to be a tumor.”

“I know this procedure.”

Such memories can help a clinician investigate an anomaly. They can also help classify the anomaly as an exception and move past it.

As experience grows, pattern recognition becomes faster. Yet professional identity can also become more tightly attached to one’s own judgment.

Contrary evidence then becomes two things at once: new information that might protect the patient, and a threat demanding the admission, “I may have mistaken the location.”

That creates cognitive dissonance.

When a belief and a fact collide, people do not invariably revise the belief. They may instead revise the meaning of the fact and make the collision disappear.

This is why the value of experience cannot be measured by case numbers alone.

Were one hundred operations one hundred occasions for learning?

Or were they one hundred reinforcements of a single conviction—“I have been right before”?

Both histories produce the same number.

Syed’s analysis also addresses the paradox that intense concentration itself can contribute to an accident.

Concentration is not the ability to see everything in one’s field of vision equally. It is the ability to narrow attention to what has been judged important and push the rest into the background. When the target is correctly identified, this ability is powerful. When the target is misidentified, skill and concentration can carry out the wrong action with increasing precision.

Whether such an attentional tunnel actually developed in this operation must await an investigation of the surgical video and interviews with those involved. But the question should not end with “How could this happen despite so much experience?” We must also ask: because experience and concentration were so strong, what method of refutation was needed to counter them?

True expertise is not the inability to be wrong. It is the capacity to protect the patient before protecting one’s experience when evidence against that experience appears.

Ending the explanation with “complacency” prepares the next accident

At the hospital’s news conference, the head of the department reportedly identified an assumption based on visual appearance as a major problem and agreed when a reporter asked whether experience had produced complacency.

Complacency offers an intuitively satisfying explanation.

A capable physician became too confident.

He discounted the warnings.

Therefore, punish that physician severely and remind everyone else to be more careful.

But this story closes too neatly.

The danger of “finding the culprit,” emphasized in Black Box Thinking, does not mean that accountability is wrong. Intention, serious rule violations, professional duties of care and responsibility for organizational design must all be examined on the evidence.

The problem arises when blaming an individual makes the explanation feel complete.

If the cause is summarized as “a complacent doctor,” the hospital can tell every other doctor not to be complacent. But confidence and assumption cannot be eliminated by an instruction to pay more attention.

The next step is not to search for a person who will remain perfectly cautious.

It is to build a system capable of stopping a person even after caution has failed.

Japan’s Ministry of Health, Labour and Welfare likewise explains that the medical accident investigation system is intended to prevent recurrence rather than pursue individual liability, and that investigations must attend to the structures that produced an accident rather than looking only at the individual. Ministry of Health, Labour and Welfare, “Q&A on the Medical Accident Investigation System,” Q24

This does not erase personal responsibility.

It means investigating personal and organizational responsibility together instead of treating them as rivals.

Patient consent is not a blank cheque written to a physician’s certainty

The case also raises a problem of informed consent.

The patient consented to surgery for the purpose of removing a tumor.

Surgery necessarily includes a risk of complications, including damage to healthy tissue. But consent is not a blank cheque authorizing the removal of anything a physician believes to be tumor.

The ethical justification of the operation depends on knowledge: the target is a tumor, and the expected benefit of removing it outweighs the risk.

As serious doubt grows over whether the target is actually a tumor, the justificatory foundation of the same physical act becomes weaker.

A negative pathology result is therefore not only diagnostic information. It is also information requiring the team to reconsider whether it may continue exercising the authority entrusted to it by the patient.

Under anesthesia, the patient cannot say, “Please stop and check again.”

She cannot see what is being removed from her body.

When the patient cannot speak, the system must doubt on her behalf.

If a serious question arises about the identity of the target in an irreversible procedure, the burden should no longer be to prove why the operation must stop. The burden should be to prove why it may continue.

Continuation should not remain the default despite doubt. Pausing should become the default until the doubt is resolved.

That may be what it means to keep the patient’s consent alive throughout the operation rather than treating it as a signature obtained beforehand.

What was missing was not information, but the authority given to information

This accident included safety-relevant information from pathological examinations.

Yet the accident was not stopped.

That fact suggests that “perform more tests” and “be more careful” are not sufficient responses.

What is needed is an epistemic brake that engages when contrary evidence appears.

At minimum, the investigation should establish:

• What anatomical evidence supported the judgment that the resected tissue was the tumor?
• When, to whom and in what words were the first and second pathology findings communicated?
• How did the surgical team share the conflict between visual assessment and pathology?
• Were the imaging, navigation data and anatomical position rechecked?
• Were options other than continuing considered?
• Who possessed the authority to pause the operation, and could team members actually voice an objection?
• Was there a predetermined procedure for a second negative finding?

These questions do not dilute the question of who was responsible.

They identify which defenses the accident passed through.

The US Agency for Healthcare Research and Quality teaches a “Two-Challenge Rule” in its medical teamwork program. When a serious safety concern has been voiced twice without resolution, the process should be stopped immediately regardless of rank. The important point is not simply that someone brave enough must speak. Once an objection meets defined conditions, it must become a stopping authority higher than individual status. AHRQ, “Two-Challenge Rule”

Kyoto University Hospital reportedly preserved the surgical video and, after the accident, began requiring confirmation by multiple people in brain-tumor resections. TV Asahi’s follow-up report

The video can become medicine’s equivalent of a black box.

Its value, however, is not to locate the scene most useful for condemning someone.

Its value is to reconstruct the sequence: where the judgment became fixed, which information arrived and why the action did not change even after that information was received.

Nor should prevention end with the abstract phrase “confirmation by multiple people.”

Who confirms?

What do they confirm?

Which result requires a pause?

Who authorizes a restart?

Where is disagreement escalated?

Only when those questions are answered does confirmation become a brake rather than a ritual.

Conclusion: a hospital able to stop even its best physician

It is not enough to summarize this medical error by saying that even an experienced physician can make a mistake.

Human beings make mistakes.

Experienced people make mistakes.

And experience does not merely reduce error; it can sometimes make an error more persuasive.

A safe organization is therefore not one that entrusts everything to its most experienced person.

It is one in which evidence can interrupt even the most experienced person’s judgment.

If surgery continued after pathology twice indicated that the tissue was not a tumor, the missing element was not information.

The information had not been given the authority to stop human certainty.

More test results alone cannot protect a patient.

A culture that welcomes contrary evidence is necessary, but it is still not enough.

An organization needs a concrete procedure that, the moment contrary evidence appears, pauses the hands, distributes authority and rebuilds the hypothesis.

Learning from failure is not only reflection after an accident.

It means creating a point inside the operation, before the failure is complete, where the team can say: “We may be wrong.”

A hospital should not merely hear warnings. It must be capable of being stopped by them.

That is the central question the investigation at Kyoto University Hospital should answer.

---

Questions raised by this case

Who should have been required to present what level of evidence before overriding two findings that the tissue was not a tumor and continuing the operation?

When the judgment of an experienced expert conflicts with a system designed to stop expert judgment, which should we trust?

How can medicine hold an individual responsible for an error while also investigating the structure capable of producing the same error again?

When a patient cannot speak during surgery, who should protect the possibility that she would want the team to stop?

---

References

• Kyoto University Hospital, “Case of resection at the wrong site due to misidentification during brain-tumor surgery”
• Kansai TV and FNN, report on the removal of tissue from a region without the tumor
• TV Asahi, “Kyoto University Hospital mistakenly removes tissue during brain-tumor surgery”
• TV Asahi, follow-up report on the experienced surgeon and complacency
• MBS News DIG, report on the mistaken removal of normal tissue
• Discover 21, Japanese edition of Matthew Syed’s Black Box Thinking
• Internet Encyclopedia of Philosophy, “Karl Popper: Philosophy of Science”
• AHRQ, “Two-Challenge Rule”
• Japan’s Ministry of Health, Labour and Welfare, “Q&A on the Medical Accident Investigation System,” Q24
• WHO, “Safe surgery: tools and resources”