The new GMC guidance on consent states that doctors should usually disclose a risk of serious harm, “however unlikely it is to occur” (23(d)). I think this is a mistake. First, it requires doctors to know all these risks, which is unrealistic. Take, for example, amoxicillin.

It carries a minuscule risk of aseptic meningitis. You can also get interstitial nephritis, seizures, vocal cord swelling and dozens of other potentially serious conditions ( https://wb.md/2F01k4C ). Should doctors know all these?
For every drug/intervention?

Second, even if doctors do know ALL the risks of serious harm, should they usually tell patients about these risks, including the risk of death, when prescribing penicillin? Even if somehow practicable, I doubt such disclosure would enhance patient autonomy.

I suspect disclosing any risk of harm, however theoretical or negligible, would often confuse and overwhelm rather than clarify, and would paint a misleading picture of the true risk of an intervention. What do you think?

@gmcuk @RCPLondon @RCSnews @lsutherlandqc @IMEweb

That should have read "any risk of *serious* harm".