The studies about regret over surgery have some problems when being used to guide changes in policy regarding when surgery is warranted. I'll explain the principle with an analogy.

Suppose you run a "hike up the Grand Canyon" program where you fly people via helicopter into the bottom of the Grand Canyon and they hike up to the top.

It's a very exclusive, limited program. It's very hard to get into. Lots of requirements. Furthermore, people think those who want to go are kind of crazy: "Who would want to climb UP the Grand Canyon?!"

But you run your program, and you have space to admit only the people who say things like, "My whole LIFE I've been DESPERATE to get to the Grand Canyon! I dream every night about hiking up it and absorbing all the majestic splendor!"

A year later you go back and ask them how it was. "Oh it was WONDERFUL!" they gush. "It was the best experience of my life!" 99% positive reports.

Well, okay then! Taking people to the Grand Canyon is great.

Fast forward a few years. Your funding is way up. People are thinking that maaaybe this isn't so crazy after all.

So you take the people who need to go, but you also walk to a bus stop outside a middle school and say, "Hey, anyone wanna hike up the Grand Canyon? Raise your hand if you want to go!"

You've got the funding, so you take everyone who raises their hand--it's only 5% of the total people so no biggie--along with everyone else who you select using the original stringent criteria.

A few years later you go back and ask everyone about their experiences. 96% positive reports! Awesome!

Grand Canyon is amazing. It's great for middle school students!

Do you see the problem?

Let's suppose you took 1000 people that year, 950 through the usual criteria, 50 from middle school bus stop hand raising.

Out of the 950 you took using the normal criterion, 940 thought it was great and 10 thought it wasn't. Out of the midde school students, 20 thought it was great and 30 thought it was awful. Overall: 96% positive.

But taking middle school students was a bad idea. You just didn't notice it was bad for them because you didn't specifically ask them. You tried to use the average of the whole population to decide how to adjust the margins.

This is exactly the problem with even the best done papers that I know about. They don't ask the question of whether relaxing criteria is a good idea--and even a more general societal tolerance of being trans without changing the medical requirements is in fact a relaxation because it changes who is willing to go through with it!

Now, the numbers are pretty high with satisfaction--usually upwards of 90%, depending on the paper--and usually the relaxations are pretty modest, not like the extreme example I gave above to illustrate the point.

But as far as I've ever seen, we haven't actually addressed the right question with research in order to guide policy in a sensible way. So we should be a little more tentative in our proclamations, and not get caught in the dichotomy between transphobes saying it's universally abusive and horrible and trans advocates saying that any questioning of surgery is transphobic.

I have no quarrel with trans people who say that they would have loved to have gender-affirming surgery earlier--this isn't the kind of thing that it's likely someone would get wrong. I wish they could have had surgery earlier, but I lack a time machine. However, the question is then: with what confidence can you identify that this other person is the same as you were? This is what is very important to get right.