In the first of a new occasional series, Tomasine Kushner and David Thomasma guide us through a medical student's dilemma

Although there is a lot of literature on ethics, medical trainees do not think that they are being adequately prepared for the daily struggles they will face when they qualify. "What do I do now?" is a column devoted to breaking what, until now, has been a code of silence surrounding the most troubling issues faced by trainees at all levels as they try to balance learning medicine, performing procedures, and interacting with patients and colleagues.

We invite readers to submit their own dilemmas for possible presentation and discussion. In all cases scenarios are presented anonymously to prevent identification of individuals and institutions involved.

Thomasine Kushner, University of California, Berkeley, and David C Thomasma, Loyola University Chicago Stritch School of Medicine, are coeditors of the forthcoming book Ward Ethics, published by Cambridge University Press, from which some of the cases and commentaries in this column will be drawn.

Commentary:

When I was a medical resident I prescribed an overdose of an anticoagulant for an elderly man who being treated for deep vein thrombosis. At that time, the drug Dicumarol was being replaced by Coumadin (warfarin sodium - an anticoagulant), which was 10 times more powerful. In writing the prescription I used the more up to date medication but inadvertently put the decimal point in the old fashioned location. Within a few hours the patient was producing beet juice urine. I was horrified. I had just turned the patient's illness from one that was potentially life threatening to one that was imminently life threatening.

I told the patient what had happened, but I cannot take credit for lofty virtue in doing so. I had no choice. We were in this together. I spent the rest of the day and the entire night by his bedside following his heart rate and blood pressure. I gave him vitamin K, checked his urine and stools, and stuck the poor man's fingers to draw haematocrits - all this after I had typed and crossmatched his blood, and transfused him with three units. He seemed to be bleeding everywhere I looked.

Throughout this adventure the patient kept trying to console me as much as I was trying to administer to him. Every time I had to do a painful pinprick of his finger to get blood, I must have conveyed my own pain so vividly that the patient kept reassuring me that it wasn't so bad. In the early hours of the morning he began to produce a more elegant vin rosé, and we both took the occasion to offer it to each other in celebration.

I was struck by how much the patient continued to trust me even after I had given him good reason not to. Somehow, the fact that I promptly confessed my error, promptly tried to correct it, kept close watch, and thereby made clear that I would do my best to protect him - all this reassured him. He could see we were in this together.

The outcome made it easier, of course. We were lucky. He came away intact. And I came away wiser.

This, in fact, is how most doctors absorb their most powerful and unforgettable lessons: from their mistakes. So that today, when I hear on rounds about a student's or resident's mishap, I'm quick to confess my own failings. "If it's true you learn from your mistakes," I say to the miserable soul, "someday I'll know everything."

Commentary

A patient's medical record is sacrosanct in the sense that it must be consistently maintained with scrupulous attention to completeness and accuracy. Under no circumstances, and certainly not for the purposes of disguising someone's mistake, should any misleading information be placed in the record. Writing a note that suggests that a patient experienced an allergic reaction to a particular medication when that is not the case cannot be justified on any grounds. To do so would deprive the patient of that medication in the future when it may be critical to his or her health.

If the treatment of the patient became an issue in subsequent litigation, the person who entered that note in the record could be called to testify about the incident. That individual would have to choose whether to acknowledge the inaccuracy of the note or stand by the accuracy of the note as a factual representation of what happened to the patient, and hence commit perjury.

Medicine is practised by fallible human beings. Mistakes are made, and when they are discovered and have adversely affected a patient they must be acknowledged. The place of truth telling in medical ethics has been encumbered by what might be referred to as "the therapeutic privilege." Technically, that phrase refers to one of only two recognised exceptions to the general rule that the patient's informed con- sent must be obtained to a procedure with any risk that is not de minumus. If, in the doctor's judgment, disclosure of the patient's condition is likely to create an unreasonable risk of serious harm to the patient the doctor may invoke the therapeutic privilege and withhold that information. Except in these exceedingly rare circumstances, the general rule of medical ethics and medical law is that a doctor has a duty to disclose to the patient all information that is necessary for an informed decision to be made about treatment options.

Historically, something like the therapeutic privilege was used by many doctors when they withheld the diagnosis of a terminal or life threatening condition from patients. Unlike the confidentiality of patient information, which has roots running deep into the Hippocratic medical corpus, truth telling as a general principle of medical ethics was not recognised until late in the 20th century. This strikes me as a curious artifact of the history of medicine if we are to think of the doctor and patient relationship as fiduciary in nature. A fiduciary is one who owes another the duties of good faith, trust, and confidence. It is by definition, therefore, inconceivable that you can discharge the responsibilities of a fiduciary while at the same time withholding from the person to whom those responsibilities run information that bears directly and substantially on that person.