Informed consent is a conversation. The form is only the receipt.

The consent form usually arrives on a clipboard about twenty minutes before an operation, carried by whoever happens to be free. The patient is gowned, cold, and has an intravenous line in the back of one hand. Somebody says the words “bleeding, infection, injury to surrounding structures,” a signature is collected, and the chart is complete. Everyone in the building has watched this happen. A fair number of us have done it.

That ritual has almost nothing to do with informed consent. It is a clerical act performed in the costume of a clinical one, and when it fails, it fails expensively — not because the paperwork was missing, but because the paperwork was the only thing anybody could produce.

What consent actually requires

Strip away the jurisdictional variation and the doctrine is not complicated. A patient can give informed consent when four things are true. They have decision-making capacity for the decision in front of them. They are acting voluntarily. They have been told what is wrong, what is proposed, what the material risks and expected benefits are, and what the reasonable alternatives are — including the alternative of doing nothing. And they have had a real opportunity to ask questions and get answers.

Where jurisdictions differ is in how they measure the third element. Some states judge disclosure against what a reasonable physician in the same specialty would have disclosed. Others — a growing group — judge it against what a reasonable patient would want to know before deciding. That second standard is the one that catches clinicians off guard, because it does not care what is customary in your department. It cares what mattered to the person on the table. A surgeon who omits a one-in-two-hundred risk of permanent numbness because it is “not a big complication” may be perfectly aligned with local practice and still be wrong, if the patient was a cellist.

A signature documents that a form was signed. It does not document that a decision was made.

Where the process actually breaks

In the cases I have reviewed, the failure is almost never a refusal to disclose. It is structural. Four patterns account for most of it.

Timing. The conversation happens after the patient is committed — after the date is booked, the childcare is arranged, the pre-op fasting is done. A patient who has already rearranged their life around a procedure is not in a strong position to hear that watchful waiting was reasonable. Consent obtained in the holding area is technically consent and practically theatre.

Delegation. The person who understands the operation is not the person having the conversation. There is nothing wrong with a trainee or an advanced practice colleague participating; there is a great deal wrong with a discussion of surgical alternatives conducted by someone who has never chosen between them.

Framing. Numbers do enormous work here, and most of us present them badly. “This cuts your risk by half” and “this takes your risk from two in a hundred to one in a hundred” describe the same finding and produce different decisions. Relative risk persuades; absolute risk informs. If you would not accept the framing from a pharmaceutical representative, do not use it on a patient.

Comprehension. We hand people documents written several grade levels above the average adult reading ability, in a second language for a meaningful share of patients, and treat the signature as evidence of understanding. It is evidence of a signature.

The single most useful thing you can do

Ask the patient to tell you the plan back, in their own words. Teach-back takes ninety seconds and does more to protect a patient — and, incidentally, a clinician — than any additional paragraph of boilerplate.

I ask three questions. What are we planning to do, and why? What are the two or three things most likely to go wrong? What happens if we decide to wait? When the answers come back garbled, I have learned something important before the incision rather than after it. When they come back clear, I have a note worth writing.

Capacity is decision-specific, and it moves

Capacity is not a status a patient holds or lacks globally. It is specific to the decision and to the moment. A patient with early dementia may lack capacity to manage a portfolio and retain ample capacity to decide about a hernia repair. A patient who was clear at eight in the morning may not be at four in the afternoon after sedation, pain and a night without sleep.

When capacity is genuinely in question, document the assessment rather than the conclusion: what you asked, what the patient said, whether they could describe the choice and its consequences and give a reason for their preference that was consistent over the conversation. “Patient lacks capacity” is an assertion. The exchange underneath it is a record.

And when you turn to a surrogate, remember what you are asking for. The question is not what the surrogate wants. It is what the patient would have chosen. Families are often relieved to be told that, because it lifts a weight they should never have been carrying.

Consent as a clinical act

The best argument for doing this properly is not defensive. Patients who understand what is about to happen to them are calmer before surgery, more accurate in reporting complications afterwards, and more forgiving when an outcome falls short of what they hoped for — because a known risk that occurs is a different experience from a surprise that arrives.

But the defensive argument is real too, and it is worth stating plainly to anyone who thinks a signed form ends the matter. In most jurisdictions that form is evidence, not proof. It establishes that a document existed. It is routinely outweighed by testimony about what was said, and the note is the only version of that testimony that was written down while anyone still remembered it accurately.

Which is the whole point. Sign the form, by all means. Then have the conversation the form is supposed to represent, and write down that it happened.