Complaints against doctors feature communication more than anything else, which is one reason why communication skills have become universal in medical education. Unfortunately we still have some way to go—as this anecdote shows.

A close friend has just been in for an operation as a day case. He’s a diabetic and has got very skilled at adjusting his blood sugar. He decided to set it higher than usual to avoid a hypo during the operation. During the operation his blood sugar went very high, and my friend believes that “these doctors know much less about controlling blood sugar than I do.” Perhaps as a result of his high blood sugar or other biochemical changes he had some sort of arrhythmia.

The operation went well apart from some minor bleeding afterwards, but because of the arrhythmia the doctors decided to keep my friend in overnight. This made him despondent.

In the evening a tired looking young doctor came to see him and said that he might have had a heart attack during the operation. They had done an ECG and another test, and my friend probably hadn’t had a heart attack, but might have done. Another test would show by the morning whether it was a heart attack.

My friend, perhaps through prolonged contact with me, has learnt not to pay too much attention to what doctors say. He was more distressed by not being able to get home than by the possibility that he might have had a heart attack, and his guts told him that he hadn’t had one anyway.

His partner, however, listened closely to what the doctor said—as, indeed, would most people. Heart attacks—unlike minor operations—may mean death and disability. She rang me, and I said that I thought that it was unlikely to be a heart attack. Nevertheless, she didn’t sleep that night, wondering what news the morning would bring.

The remarkable thing is that the morning brought no news. “Feeling alright?” asked the consultant surgeon.

“Fine,” said my friend.

“Off you go then,” said the surgeon, and my friend was out of the hospital as fast as he could.

So did he have a heart attack? Probably not. And somebody more impressed by doctors and less keen to get out of the hospital might have asked, but it is a failure that the possibility can be raised and then forgotten.

I can imagine what happened. The arrhythmia may have been linked to his blood biochemistry not being well controlled. Whatever caused it the anaesthetist asked the junior surgeon, somebody keener on and more knowledgeable about cutting than heart problems, to check things out “just to be sure.” He couldn’t see anything on the ECG, but sent blood for cardiac markers. In order to be fully open with my friend he told him that he might have had a heart attack but probably hadn’t and that results of the blood test in the morning would show. His shift then ended, the blood test results might have come back and nobody had looked at them—or perhaps they hadn’t come back. So the possibility of a heart attack is forgotten by the medical team and left hanging.

Like most medical failures this is not a failure of an individual but a failure of the system—and it’s the result of a series of failures, including the failure of the patient to ask if he’d had a heart attack. Like most system failures it hasn’t resulted in tragedy, although my friend’s partner had an anxious sleepless night, but it could have done. If my friend had dropped dead of a heart attack on his way home this would be a very different story. Communication depends not only on individuals communicating well but on the whole system working without fault, and communicating well is just as important as doing the operation well.

Richard Smith was the editor of the BMJ until 2004 and is director of the United Health Group’s chronic disease initiative.