Thread by @iwashyna

This proposed “do + report back” is literally the definition of research — changing care of a patient to create generalizable knowledge rather than focused on that specific patient—and is frankly unethical and probably illegal

#HardPass unless there is informed consent involved

Because this involves me being in a twitter fight with someone I deeply respect, let me be a little clearer

If the tweet were merely “optimize each of your patients’ care”, I would love it, and indeed I try to do that everyday

But the first sentence of the tweet proposes manipulating the care of the patient and then reporting the results of that manipulation

I think it can be read as saying the point of the manipulation to the next 5-10 patients is to learn from them

I think it can reasonably be read that the point of changing the care is to see their physiologic response and aggregate that physiologic knowledge — “and report back” adding an explicit goal of generalizable knowledge

And that is what crosses the line into research for me

And research needs fundamentally community oversight

It is entirely likely that is not the intent; god knows my tweets are full of typos and mis-statements and just plain wrong ideas + assertions

(The original author, blanked out to focus on the words nor the man, is spectacularly thoughtful about the care of individual patients)

I hope I will be forgiven for my many errors, and if the photo’d tweet is clarified to mean something different, I will absolutely ask forgiveness for overreacting and delete my objections

But until it is, I fear that the reading of it as proposing “changing the care of patients in order to generate generalizable knowledge” is tenable, and I feel obliged to maintain an objection to that being done without research oversight

Epilogue: I have been persuaded the post in question was really just an exhortation to go to the bedside of our patients, and share our general experiences

I over-reacted and I apologized to the author

To all of us, though: please do not experiment without consent + oversight