The 96 Per Cent Problem: Why Doctors Do Not Report Errors

Somewhere between 50 and 96 per cent of adverse events in healthcare go unreported. That range is itself remarkable. We do not know the actual number because the system designed to capture errors is so distrusted that most clinicians simply do not use it.

In no other high-risk industry would this be considered acceptable. In aviation, confidential reporting systems capture virtually every near miss, every procedural deviation, every moment where something almost went wrong. Line operations safety audits find, on average, two threats and two errors per routine flight. Not because pilots are reckless. Because humans operating complex systems make errors, and the system is designed to catch them.

In medicine, the system is designed to find someone to blame.

The cost of silence

The Institute of Medicine estimated that medical errors cause between 44,000 and 98,000 deaths per year in the United States. That number is more than two decades old. Subsequent research has suggested the true figure is significantly higher.

But the response to this has not been to build better reporting systems. It has been to add more regulation, more mandatory reporting, more individual accountability. And the result is predictable. Helmreich, writing in the BMJ in 2000, found that 30 per cent of doctors and nurses working in intensive care units denied ever committing an error. Not rarely. Never.

That is not a reflection of exceptional competence. It is a reflection of a culture where admitting error is professionally dangerous.

A doctor reports an adverse event and the response is an investigation. Sometimes a formal complaint. Sometimes a tribunal hearing that stretches over years. The information does not feed quietly into a learning system. It feeds into a process that assigns fault. The doctor’s name may end up in a public finding.

Compare this with what happens when a pilot files an airspace incursion report in Australia. Air Services Australia makes a phone call. Not an accusation. A conversation. What happened, what were the contributing factors, what could the system do differently. No sanction. No career consequences. The data goes into a database that makes the system safer for everyone.

The same error, different outcomes

This is the part that most people outside medicine do not understand. The same clinical error, in different hospitals, under different investigators, can produce outcomes ranging from a quiet learning conversation to a career-ending finding. There is no consistent framework for distinguishing between an honest mistake, a lapse in judgement, and genuine recklessness.

Aviation solved this problem decades ago with what James Reason called “just culture.” The framework draws a clear line. Human error gets a system response (what failed around this person?). At-risk behaviour gets coaching. Reckless behaviour gets discipline. A pilot who reports an honest error is protected. A pilot who was intoxicated is not.

Because the line is clear, pilots report freely. The data flows. The system improves.

Medicine has no equivalent that is consistently applied. And so the silence continues.

Who actually gets hurt

The assumption behind punitive reporting systems is that accountability makes the system safer. It does not. It makes the system quieter. Every unreported near miss is a signal the system never receives. Every doctor who stays silent about an error they made is a learning opportunity that disappears. The next clinician in the same situation, facing the same systemic pressures (fatigue, distraction, ambiguous protocols, poor communication), has no warning that this particular combination of circumstances nearly caused harm before.

Anaesthetists understand this better than most. The operating theatre is the closest medical environment to a cockpit: real-time monitoring, checklists, managing emergencies under pressure, working in a system where small errors compound quickly. The specialty has some of the lowest complication rates in medicine, and that correlation between systems thinking and safety outcomes is not a coincidence.

But even anaesthetists work within a broader medical culture that, when something goes wrong, asks “who” before it asks “why.”

What would need to change

The answer is structurally simple. Separate learning from punishment. Create confidential incident reporting systems that do not feed into the same management structure that conducts disciplinary proceedings. Train human factors and error management from medical school, not as a box-ticking exercise in hospital orientation. Accept that the doctor who made the error is usually the second victim, not the first cause.

Aviation implemented these changes decades ago. Australia was the first country in the world to formally apply Reason’s Swiss cheese model in accident investigation. The evidence base exists. The frameworks exist.

I wrote about the full comparison between aviation and medical safety culture, and what a just culture framework would look like in practice, here: What Medicine Could Learn from Aviation’s Safety Culture.

The question is not whether these systems work. They do. The question is whether medicine has the institutional courage to stop looking for someone to blame.

Dr Adam Hill is a specialist anaesthetist (FANZCA) and instrument-rated pilot practising in Sydney and regional NSW. He flies his own aircraft to deliver specialist anaesthetic services to regional communities. Read more at dradamhill.com.au.