It was actually even worse than that. The question was "have you ever
thought about suicide". You didn't even have to have made the most half-assed attempt, just have thought about it once or twice and that's all it took to get included in the stat.
Also iirc the survey was anonymous, self-reporting and online with no records of who took it, so it could be taken as many times as you wanted. There's a pretty high likelihood that some dedicated troons filled it out a couple of dozen times to inflate the numbers and make sure their "vulnerability" was reflected in the results.
But like, also, find me a study that trans activists love/use to support their bullshit that doesn't have massive methodological issues, logical fallacies, confirmation bias, sampling issues, etc... you can't, because all of them do.