You’re in a pre-op bay with an IV already in your arm. Someone you’ve never met hands you a clipboard with three pages of small type, says the surgeon is running behind, and points to a line at the bottom.
You sign. Almost everybody signs.
That moment is the part of informed consent people remember, and it’s the part that counts for the least. The signature is paperwork. The decision was supposed to happen earlier, in a conversation you could follow and interrupt. Consent law in the United States is mostly state law, and the standards differ from one state to the next, but the thing being protected is the same everywhere: your ability to understand what’s about to happen to you and to say yes or no on purpose.
Below, we look at what a clinician is supposed to tell you, when a signature is required, and what to do when the form shows up ten minutes before the procedure.
The form is a receipt, not the decision
Hospitals that bill Medicare have to keep that signed page. The federal medical records rule for hospitals, at 42 CFR 482.24, requires the chart to contain “Properly executed informed consent forms for procedures and treatments specified by the medical staff” or by federal or state law.
Read that wording again. The rule is about what the record must hold. It never says the form is the consent.
The American Medical Association’s Code of Medical Ethics treats consent as a sequence, not a document. Opinion 2.1.1 asks the physician to check that you can understand the information and decide freely, present it accurately and in the way you prefer to receive it, get your decision, and then write the conversation and the decision into the record. If there’s a written form, it goes in the chart too. It’s the last step.
A 2020 systematic review of informed consent in neurosurgery, published in Neurosurgical Focus, found that a failed consent process shows up repeatedly in malpractice claims. The dispute in those cases is rarely whether a signature exists. It’s whether anyone had the conversation.
What a clinician has to explain
MedlinePlus, the consumer health service of the National Library of Medicine, lists what a provider is supposed to cover before you agree to treatment. Strip it down and it answers four questions.
What’s wrong, and why this? The health problem you have, and the reason this particular treatment is being recommended for it.
What will happen to me? What the treatment involves, in enough detail that you could describe it to someone else.
What can go wrong, and how often? Not just a list of risks but how likely each one is. MedlinePlus also says this has to include side effects and complications that may not show up until later, which is the part most people never hear.
That word “likely” is doing a lot of work. “Rare” and “uncommon” mean different things to different clinicians, and they mean almost nothing to a patient. Ask for a number out of 100. If the honest answer is that nobody knows the number for someone with your particular history, that’s useful too, and a good clinician will say it.
What else could I do, and what if I wait? Other options, how likely this one is to work, whether the decision is urgent or can sit for a while, and what happens if you delay.
The AMA frames the same ground slightly differently in Opinion 2.1.1: the diagnosis when it’s known, the nature and purpose of what’s being recommended, and the burdens, risks and expected benefits of every option, including the option of forgoing treatment entirely. Doing nothing is one of the choices a physician is supposed to put on the table.
Who decides how much is enough
There’s no single national answer, and this is where state law does the work. The StatPearls clinical reference on the National Center for Biotechnology Information’s Bookshelf describes three different legal yardsticks for how much a clinician must disclose.
One asks what a typical clinician in that specialty would normally say. A second asks what an average patient would need to know to decide. A third asks what this specific patient needs, given what the clinician knows about them. According to StatPearls, many states use the average-patient standard, and the state you’re treated in determines which one applies.
MedlinePlus puts the practical consequence bluntly on its patient rights page: patient rights differ from state to state. Some are federal, like getting copies of your records. Many are not.
Consent is also not the same thing as shared decision making. Consent is a threshold you have to clear before treatment starts. Shared decision making, which MedlinePlus describes as providers and patients working out together how to test for and treat a problem, is a way of reaching the decision in the first place, and it includes the option to wait.
When your signature is required
Not every treatment needs a form. For a lot of care, a provider explains what they want to do and asks if you agree, and that verbal yes is real consent.
MedlinePlus says written consent is generally needed for surgery, including surgery done outside a hospital. It’s needed for complex tests and procedures, with endoscopy and a liver needle biopsy given as examples. It’s needed for cancer radiation and chemotherapy, and for high-risk drug treatments, where MedlinePlus names opioid therapy and corticosteroids.
Two items on that list surprise people. MedlinePlus says most vaccines require written consent. And it says some blood tests do, including genetic testing, while noting that most states have dropped the separate written consent requirement that used to apply to HIV testing.
One more thing worth knowing: even where no form is required, MedlinePlus says you should still be told what test is being run and why. Its examples include PSA tests, Pap tests and mammograms. If a tube of blood goes off for something you weren’t told about, that’s a gap.
Research consent is a separate and stricter system, and it’s easy to confuse the two. StatPearls describes valid research consent as resting on three legs: disclosure of the information, the competency of the patient or surrogate, and the voluntary nature of the decision. Research consent also runs through an institutional review board, which can waive consent only under narrow conditions, mainly when the risk is minimal, as in a records review. If someone asks you to join a study, that’s a different document and a different decision from agreeing to be treated.
Why so many people sign without understanding
The consent process has a reading problem, and researchers have measured it.
A study published in Clinical and Translational Science in 2015 collected the procedure consent forms used by every surgical inpatient hospital in Rhode Island and ran them through seven readability formulas. The forms came out at roughly a 15th-grade level, meaning about three years of college. The authors noted that the average American adult reads at about an 8th-grade level and that the AMA and the National Institutes of Health suggest patient materials sit at a 6th-grade level.
That gap is not a rounding error. It’s seven to nine grade levels.
A 2024 analysis in the Journal of Surgical Research looked at 386 surgical clinical trial consent forms posted on ClinicalTrials.gov. The average form ran 2,626 words, which the authors calculated at roughly 13 minutes of reading at a normal pace. Only 4% hit the recommended reading level.
Comprehension follows. In a study of elective orthopedic surgery patients published in the Journal of Bone and Joint Surgery in 2011, 98 people took a quiz immediately after their surgeon had discussed risks and benefits and the entire form had been read and explained to them. They averaged 70.7% correct. At the first visit after surgery, no more than two weeks later, the average had fallen to 59.5%. Older patients and patients without a college education scored worse.
The 2020 neurosurgery review found the same pattern across studies and added a detail that matters: recall was weakest for risks and for alternatives. Those are exactly the two things a consent conversation exists to convey.
So if you’ve ever walked out of an appointment unsure what you just agreed to, you are the normal case, not the failure.
The same review found that the fixes are unglamorous and they work. What improved patient knowledge was combining formats rather than talking alone: a list of questions handed to the patient before the discussion, illustrations, interactive websites, and a consent form with specific points the surgeon has to check off as each one gets covered. None of that is exotic. It’s an admission that a five-minute speech is a bad way to move complicated information into a nervous person’s head.
Teach-back, and why you should ask for it
There’s a cheap fix, and it has a name. The Agency for Healthcare Research and Quality, the federal agency that studies health care quality, publishes it as Tool 5 in its Health Literacy Universal Precautions Toolkit: teach-back.
The clinician explains something, then asks you to say it back in your own words. Not to test you. To test the explanation. AHRQ is specific that the question should be framed as a check on the clinician’s clarity, and that “Does that make sense?” is a bad question because people say yes either way. AHRQ also recommends chunking information into small pieces and checking after each one instead of saving it all for the end. If you repeat the clinician’s exact phrasing back, AHRQ warns that’s a sign you may not have understood it.
Does adding this step make patients feel rushed or quizzed? A multicenter randomized trial in the Veterans Affairs system, published in the Journal of Patient Safety in 2014, tested it on 575 surgical patients. Half could not sign until they’d satisfactorily restated the key elements. Satisfaction, ease of understanding and perceived time to decide were the same in both groups. One thing did improve: 80% of the teach-back group strongly agreed they’d received the right amount of information about alternatives to surgery, compared with 69% of the others.
You don’t need the hospital to adopt a program. You can run teach-back yourself: “Let me say back what I think you told me, and you tell me what I got wrong.”
Capacity, surrogates, and the emergency exception
Consent requires that you be able to decide. StatPearls defines that ability, called capacity, as being able to take in and hold the information, weigh the options, and communicate a choice. It makes a point worth repeating: cognitive impairment, mental illness or serious illness on their own do not strip a person of the right to consent, as long as those abilities are intact. Capacity is also decision-specific. Someone may be able to agree to a flu shot and not to a bowel resection.
When a person genuinely can’t decide, because of advanced dementia or a coma, MedlinePlus says providers turn to a surrogate decision-maker. If no one has been named in advance, StatPearls explains that state law sets an order of who gets asked, and that a court can appoint a guardian when no suitable person exists.
Minors are handled differently. StatPearls describes parents or guardians giving informed permission rather than consent, with legally emancipated minors able to consent for themselves. Its examples of emancipation include being married, serving in the military, or being financially independent. State laws vary, and that variation is wide.
Capacity also isn’t something the person with the clipboard decides on their own. StatPearls notes that a psychiatrist may be asked to evaluate it when it’s genuinely in doubt. If a family member is told that a relative “can’t consent” and nobody has explained how that conclusion was reached, that’s a fair question to ask out loud.
Then there’s the exception that swallows the clipboard. MedlinePlus says informed consent is not required in an emergency when delaying treatment would be dangerous. The AMA puts it the same way in Opinion 2.1.1: when a decision is urgent, the patient can’t take part and no surrogate is available, physicians “may initiate treatment without prior informed consent,” and should inform the patient or surrogate as soon as they can and get consent for continued care.
This is why nobody asks an unconscious trauma patient to sign anything. It’s also why the exception is narrow. Running late is not an emergency.
What to do when the form arrives ten minutes before
This is a sequence built for the situation most people face, which is a form appearing after the gown is on.
First, ask four questions. What exactly are you doing, in one sentence? What are the two most likely things that could go wrong, and how often do they happen? What happens if I wait a month? What else could be done instead?
If the person holding the clipboard can’t answer those, they’re the wrong person. That’s not a complaint about them. Admitting staff and pre-op nurses are often handed the form to collect, not to explain.
Second, ask for the person doing the procedure. You’re entitled to hear it from them. The AMA’s patient rights opinion includes the right to ask questions when you don’t understand a recommendation and to get answers.
Third, run your own teach-back. Say the plan back in your own words and let them correct you. Thirty seconds, and it catches most misunderstandings.
Speak up, specifically, if any of these are true. You’ve already been given sedation and you’re being asked to sign now. The procedure on the form isn’t the one you discussed, or it names a different side or a different joint. You were told about one option and the form lists several. Nobody has mentioned a single risk. Any of those is a reason to say the words “I want to wait until I understand this,” and a competent team will stop. If you’re in an emergency, this calculus changes, and letting them work is usually the right call.
Bring someone if you can. A second set of ears in the room is the simplest protection available, and it costs nothing.
The most useful thing to take from this
If you remember one line, make it this: the signature is not the consent, so a form you didn’t understand is a problem you can still fix before the procedure starts, and saying so out loud is the whole fix.
Consent is also not permanent. MedlinePlus states plainly on its patient rights page that you can change your mind at any time. Agreeing to a plan in September does not lock you into it in November. If something you learn later changes the math, the conversation reopens, and your clinician’s job is to have it again.
Sources
- MedlinePlus, National Library of Medicine. “Informed consent – adults.” medlineplus.gov
- MedlinePlus, National Library of Medicine. “Patient Rights.” medlineplus.gov
- MedlinePlus, National Library of Medicine. “Shared decision making.” medlineplus.gov
- StatPearls, NCBI Bookshelf, National Library of Medicine. “Informed Consent.” ncbi.nlm.nih.gov
- American Medical Association. “Informed Consent: Code of Medical Ethics Opinion 2.1.1.” code-medical-ethics.ama-assn.org
- American Medical Association. “Patient Rights: Code of Medical Ethics Opinion 1.1.3.” code-medical-ethics.ama-assn.org
- Agency for Healthcare Research and Quality. “Use the Teach-Back Method: Tool 5, Health Literacy Universal Precautions Toolkit.” ahrq.gov
- Electronic Code of Federal Regulations. “42 CFR 482.24, Condition of Participation: Medical Records Services.” ecfr.gov
- Electronic Code of Federal Regulations. “42 CFR 482.13, Condition of Participation: Patient's Rights.” ecfr.gov
- PubMed. “Eltorai AEM et al. "Readability of Invasive Procedure Consent Forms." Clinical and Translational Science, 2015. PubMed. DOI: 10.1111/cts.12364.” doi.org
- PubMed. “Crepeau AE et al. "Prospective evaluation of patient comprehension of informed consent." Journal of Bone and Joint Surgery (American), 2011. PubMed. DOI: 10.2106/JBJS.J.01325.” doi.org
- PubMed. “Prochazka AV et al. "Patient perceptions of surgical informed consent: is repeat back helpful or harmful?" Journal of Patient Safety, 2014. PubMed. DOI: 10.1097/PTS.0b013e3182a00317.” doi.org
- PubMed. “Karimi AH et al. "Assessing the Readability of Clinical Trial Consent Forms for Surgical Specialties." Journal of Surgical Research, 2024. PubMed. DOI: 10.1016/j.jss.2024.01.045.” doi.org
- PubMed. “Shlobin NA et al. "Informed consent in neurosurgery: a systematic review." Neurosurgical Focus, 2020. PubMed. DOI: 10.3171/2020.8.FOCUS20611.” doi.org