[Houshalter]

The hospital allowed them to have pencils and apparently had no issue with it. Presumably patients fill out forms with pencils all the time. Even the IRB failed to complain or notice this "danger". They never complained that pencils were too dangerous, just that it didn't meet the arbitrary signature requirement that it be a pen.

And you say this is the most ludicrous requirement, but it's actually the least. After all it's not entirely the IRB's fault, they didn't ban pens in the hospital.

But it demonstrates exactly what is wrong with bureaucracy. Maybe one bureaucracy with one set of arbitrary rules could be tolerated. But once you have two interacting they can create contradicting and incompatible rules. One bureaucracy bans pens, the other bans pencils, and you end up with a world with no writing implements at all.

I'm sure the regulation sounded entirely reasonable at the time. Someone had to put together a set of rules on how to get consent. And they think pens are supposed to be the proper way of writing a signature, because banks require you to sign checks in pen. Because long ago people would erase checks written in pencil and commit check fraud. At no point did they ever consider the history of that rule or that it doesn't make much sense applied to consent forms. At no point did they ever consider that there might be a hospital somewhere that bans pens for whatever reason.

And I can't blame them, why would they? You can't anticipate every possible edge case. This is a website made up of programmers, we should know that better than anyone. And yet at no point did anyone with common sense come along and make an exemption for that rule. When rules become fixed and inflexible they do things that weren't intended by the rule writers.

And every single thing in the article is like that. Some rule that might not be so bad or make sense in isolation. But combine it with 10,000 other rules, and total weight becomes overwhelming. And as a result you get a bunch of people misdiagnosed with bipolar disorder, and god knows what else since doctors can't do the research necessary to find out.

[euyyn]

> At no point did they ever consider the history of that rule or that it doesn't make much sense applied to consent forms.

Or maybe they did?

> At no point did they ever consider that there might be a hospital somewhere that bans pens for whatever reason.

IRBs are part of each research institution. The unspoken implication in OP's story is that the psychiatry department of his hospital had never conducted research with humans before him. At least not any research that required consent. If they had, they would have faced the need for pens. "Requiring IRB review and a consent form" is almost the opposite of an edge case for a research institution.

> Some rule that might not be so bad or make sense in isolation. But combine it with 10,000 other rules, and total weight becomes overwhelming.

I can't really feel that steps like being forced to blind your data, or to store that data somewhere it can't be read by random people, are overwhelming, nor ludicrous, nor indefensible. I understand that learning by hitting walls is frustrating, but those things are proper experimental procedure. Maybe his professors should have had a lecture about them (mine did). Or maybe it was the purpose of that video that he thought was a waste of his time because he's not a Nazi.

> Even the IRB failed to complain or notice this "danger". They never complained that pencils were too dangerous

If I'm reviewing that application and on the risks section I see "paper cuts lol", and the applicant then asks me to allow pencil signatures because pens at his department are a risk, I would conclude he's not taking any of it seriously enough.

[yebyen]

Just playing devil's advocate, what is the possible danger of accepting a consent form with a signature in pencil? Is someone going to erase a signature from the consent form? (And how exactly is that a potential risk or danger?)

I'm pretty sure that it's not any harder to forge a signature when it's made using pen or pencil. It's just a mark that says "I agree." It's not a bank form where the validity of the amount tendered could be called into question.

[euyyn]

I could imagine this:

"Oops, I forgot about presenting my subjects with the consent form and now I've got 100 results collected during a whole year of work.

I definitely don't want to throw all of that work away... But if I contact them and make them sign the form now using pencil, I could erase the date and change it to each subject's day of experiment. Nobody would notice."

Pen raising that barrier. I agree that against an adversary fully committed to fraud it doesn't matter. But that threat model is exceptionally rare, and an IRB can do little about it anyway (e.g. the whole application could be made up, the experiment consisting of something completely different). The realistic threat they fight against is experimenters that don't know better, and think there'd be no harm in their actions.

And this is just me coming up with a possibility in a few minutes. Reality tends to be richer than a single person's imagination.

[yebyen]

Some researcher who paid $40 to 100 people over the course of a year is going to go back and contact all of those same people again, to sign a release form, and they all have to use pencils... and nobody is going to call him out on it and ask why is it important that I sign this form with a pencil?

The fact that I actually agree with you, and think this might work in fact inside of any hospital in the US, says more about the byzantine complexity and scary state of our medical system than anything else.

Because I could totally see 100 people being told they have to sign a thing, while in a hospital, with absolutely no good reason given, ... and all 100 just signing it after some variable x*N amount of huffing, then get on with their day.