The Lie in the File

Three years ago a car pulled up outside my practice and a woman I knew got out of the driver’s seat, which was wrong, because her husband never let anyone else drive. He was in the passenger seat. He had been driving when it started, she told me; he had suddenly stopped making sense, and she had somehow got the car stopped and got him here. She was scared. I went out to him, a man I had known for years. A patient at times, but more than that, a friend. He looked at me the way you look at a stranger who is standing slightly too close. He did not recognise me, nor the clinic. Somewhere behind his eyes something had come loose, and the man I knew was not answering.

There was no reasoning with him. He felt well, he said, though he was plainly confused and failing, and he refused to enter the clinic. What he wanted was to get back in his car and drive, he knew not where. I had an ambulance called, and when the crew arrived, mercifully fast, he fought us all. He is a strong man, and in his confusion he was stronger still. The paramedics reached for ketamine, a serious drug they keep for the far end of the agitation scale. Nothing happened. He was untouched by a dose that settles almost anyone. He struggled on, and only after the third injection could we get him into the ambulance. Three doses. Hold onto that detail, because it matters: a triple dose of ketamine does not happen with a dizzy spell.

Walker Briefing is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

He was admitted to a big hospital under a consultant neurologist and his team, and I went back to my clinic to worry. Several days later I learned two things. The first was that he had been discharged and was already in Melbourne, where he had hired a car and was driving. The second stunned me. His diagnosis was postural hypotension. A drop in blood pressure on standing. The terrifying thing I had witnessed had lifted for him the way a fog lifts, the tests had come back inconclusive, an MRI of the brain among them, and so with nothing concrete to write, something tidy was written instead. Postural hypotension. And with that benign label he was cleared, discharged, and sent on his way interstate.

In all my clinical career I have seen many episodes of postural hypotension, and if I can be certain of one thing it is this: whatever felled my friend that day, it was not a drop in his blood pressure. You do not need my bald head to see it; the guidelines every doctor works from say the same. A person recovering from a faint is clear-headed almost immediately, disorientated for perhaps five or ten seconds. Confusion lasting many minutes points away from a faint altogether, toward a seizure among other possibilities. An episode that runs for the better part of an hour, in a man who is awake, on his feet, fighting three grown adults and unable to recognise a friend of many years, is not what a faint looks like. It is not even close.

And there was an honest label available. Medicine has one for exactly this situation, and it is no shame to use it: cause not yet determined. We looked, we found nothing conclusive, we do not yet know. Every doctor writes it somewhere in a working life. That label is not a failure. It is the truth, holding the door open for the answer.

But the honest label has consequences, and here is where this stops being a story about one man’s file. The national standards that govern fitness to drive treat honest uncertainty with real respect: an unexplained episode of this kind sits in the same family of standards as blackouts and first seizures, six months off the road for a private driver, because the next one might arrive at a hundred kilometres an hour. An explained episode, a simple faint with an obvious trigger, carries almost no restriction at all. Do you see what the label did? Unexplained, and the safeguards wake up. Explained, and they sleep. The tidy diagnosis did not just close his file. It put him behind the wheel of a hire car in Melbourne traffic, within days, with a question mark in his brain that nobody had answered. The label was the licence.

My mind went straight to the story I could not stop writing in my head: a man who has suffered some major neurological event, discharged unexplained, and an avoidable accident with someone dead at the end of it. If that happened and I had said nothing, it would have been on me for the rest of my life. So I rang the hospital. Not to lodge a complaint. To ask a doctor’s question: what did you find, and how does it lead to this diagnosis? Walk me through it, because I saw the acute event, which nobody at your hospital did, and I cannot get there.

The junior doctor on the ward gave me a short and unenlightening conversation and grew defensive. Then her registrar rang me back, not with findings but with a complaint: the diagnosis was clear, and why was I casting doubt? She had heard my questions as an attack, which they were not. And then the consultant wrote to me. Not with the findings. Not with the reasoning. With a demand that I explain myself for calling his diagnosis into question. I have the letter still.

What I felt when I read it belongs in the evidence, so here it is. I felt insulted. Underneath the courtesies, the message I received was that a general practitioner could not be expected to know the difference, and should not presume to ask. There was no differential diagnosis on offer, no list of possibilities the GP might watch for, nothing for me to work with in the years of his care that would now be mine. And far stronger than the insult was disbelief. I was not a rival. I was offering that hospital the one piece of clinical evidence it lacked, an experienced clinician’s account of the event itself, and the institution’s whole reflex, from intern to consultant, was to defend the label rather than examine it.

The letter did not get the contrition it seemed to invite. I have been leaned on inside the healing professions often enough to know how to return serve when the stakes are a patient’s life, and I called the consultant and had a rather one-sided conversation, at the end of which he understood that no referral of mine would ever again cross his desk. He maintained to the last that his diagnosis was correct. I was sorely tempted to take it further. I let the temptation pass.

Why does a system produce that piece of theatre? I cannot see inside another man’s head, so take what follows for exactly what it is, a judgement from long experience and nothing more. Sometimes a discharge demands a diagnosis because the paperwork will not accept “unknown”, and the bureaucracy is pleased. Sometimes a clinician has been taught, without anyone ever quite saying it, that a precise-sounding label is competence and “we do not know” is failure. I have sat on the other side of this myself: asked, in my time, to write a diagnosis that did not reflect our real uncertainty, because uncertainty would read as failure. For the record, I refused, wrote the differential and the steps that would settle it in due course, and let it be challenging. A specialist may find that uncomfortable. A GP has to be comfortable with uncertainty; it is most of the job.

For all the heat in what I have just told you, nobody in this story is a villain, and the strongest evidence is how uniform the reflex was. The junior did what juniors learn to do. The registrar and the consultant defended what their training and their institution taught them to defend. Three people at three levels ran the identical programme, and that is not a coincidence of character; it is a culture. Study after study of diagnostic error finds the same engine at the bottom: premature closure, the reaching of an answer and the closing of the question, after which every new fact is bent to fit the label or brushed aside. The label acquires momentum. Each doctor who opens the file meets the verdict before they meet the patient. Challenge, which is the only medicine that cures a wrong diagnosis, gets processed as insolence instead of information. It happened to me, a colleague of forty years’ standing, doctor to doctor. Now imagine how it goes for a worried spouse.

There is a boundary here, and I will respect it. I do not know what was in their test results beyond what I was told, because they never showed me the reasoning. Three years later my offer stands: show me the workings, and whatever they reveal that day to have truly been, I will publish it as plainly as I am writing this. But that is precisely the point. A diagnosis whose evidence cannot be shown to the patient’s own doctor is not knowledge. It is a ruling, and medicine is not supposed to issue rulings.

And I have come to think this reflex is not a hospital problem at all, but what power does with questions everywhere, once being certain becomes part of the job. I sit in a Parliament where I ask institutions to show their workings for a living, and I meet the same sequence more often than I can tell you: the confident answer, the refusal of the underlying data, and, if you press, the suggestion that the question itself is the offence. A hospital defending a diagnosis and a department defending a decision are running the same programme. The file must be right, because the file is ours.

You will want to know how my friend is. He is well. Three years on, there has been no second episode, and I am more glad of that than of anything else in this story. But be careful with the comfort in it. The good outcome does not mean the label was right; nothing we ever learned says it was. It means we were lucky, or it means the true cause was something that passed, and we will never know which, because the question was closed on day one and nobody was permitted to reopen it. He is well, and his file still contains a certainty no one ever earned. Both of those things are true, and it is still puzzling, and honest medicine would have let it stay puzzling out loud.

What would the honest version have looked like? Small, and completely different. A discharge summary that said: episode of confusion, cause not established, investigations so far unremarkable, do not drive until this is resolved, GP to follow up, here is everything we found. A phone call answered with a colleague’s curiosity instead of a defended border. A system in which “we do not know yet” is a complete and honourable clinical sentence, spoken to the patient’s face, written in his record, and shared with the doctor who will carry his care. Nothing in that version needs a hero. It only needs institutions that treat a question as a gift rather than a threat, because in medicine a question is very often the last safeguard left standing between a wrong label and a harm nobody meant.

Everything I trust about healing begins at the same place: the first honest sentence. Between a doctor and a patient, between a hospital and a GP, between a government and its people. Certainty that has not been earned protects nobody. It only decides, in advance, who will not be listened to. My friend was owed that honest sentence three years ago, and so was I, and in a quieter way so is everyone whose file holds an answer that was really a full stop.

So carry this story with you to question time, next time you watch our Parliament at work, and listen the way a doctor listens. Are the questions being answered, or is the file being defended? You will recognise the reflex now, in a hospital, in a department, in a chamber. Naming it is not cynicism. Naming it is the first honest sentence, and everything that heals starts there.

Walker Briefing is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.