New mental health analysis of Tavistock pb trial

From the discussion section of the original analysis: However, some conclusions can be drawn. It is unlikely that the reported adverse events such as headaches do not relate directly to GnRHa treatment. Equally, given that there were no changes in psychological function and differences in point estimates were minimal for nearly all outcomes, it is unlikely that the treatment resulted in psychological harm.

OH! OK, I actually get why this is news now. Presumably the "no psychological harm" part of the conclusion was previously reported. That's now been undermined by the re-analysis. I see the piece was written by Hannah Barnes. I suspect it might read as inconsequential because it was very heavily edited.

And perhaps they feel that the publicity that may have been given to the earlier analysis requires them to air this corrective re-analysis,

I'm not sure there was a lot of publicity given to the earlier analysis? It's linked below and it's difficult to see how you could spin it into something attention-grabbing. But I agree with you the article is quite factual, it's just very much "in the weeds" and it's odd reading because of that.

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0243894#sec020

I can't help wondering if they can't start the 'nhs trials' without designing the trial properly which means working out how the data will be collected and reported?

I hope so! Design it properly from the start, and no monkey business.

I thought the BBC article was careful, factual, adequate. I think they have simply caught up with the idea that they have a particular duty on this issue to engage in careful, comprehensive fact-based reporting. And perhaps they feel that the publicity that may have been given to the earlier analysis requires them to air this corrective re-analysis, despite the smallness of the study, the provisional nature of its findings, the clear need for further research, about all of which the article is very clear.
The very fact that this makes the article seem somewhat inconsequential shows that they are trying to subtract the journalistic hype that has tainted earlier reporting.

It is all a bit weird. I’m not sure how you distinguish statistically between “some people got better, some people got worse, some experienced no effect” and “puberty blockers had no effect” with a sample size of 44.

But since the original study authors welcomed it, I’m willing to accept there may be something useful about this re-analysis. I don’t see how it can possibly merit a BBC headline though! I’m guessing it’s a sign that the TRAs at the BBC are absolutely desperate, because at face value I don’t think this is particularly good news for them. But they are so desperate to promote the idea that “trans kids” are a real thing that any evidence that some kids might be helped by puberty blockers is apparently headline news.

Yes, exactly. And scunnered is the word! Along with feart.

Eventually there will be some but don't hold your breath!

At the moment it's important to just recognise how little reliable data has been gathered about the effects of transition overall and especially childhood transition, and from a practical point of view to accept that gender clinicians really have been making "strong recommendations based on weak evidence" (Dr Julia Walker)

From a research point of view it's normal for there to be discussion and disagreement about how to interpret research findings. It's the attempts to cover up and silence disagreement that are most worrying.

It's not an area where definitive answers are easy to get. But if I was funding studies I would put much more emphasis on physical health outcomes which up to now seem quite neglected.

So, ultimately, we are still treating children with medications based on fuck all evidence that the benefits outweigh the risks, yes?

I'm scunnered by this.

Thanks for the clarification @MumOfYoungTransAdult and @NotBadConsidering.
Having read so many dodgy "research papers" on pbs I sometimes get them mixed up. I'm still trying to find any that stand up to rigorous scrutiny but haven't found any yet, so if anyone knows of any I'd love to read them.

That's not what this paper is, though.

Keeping this in the public eye is a good thing, and reporting truthfully is a good thing.

The BBC summary of the paper's results and interpretation was very clear and pretty balanced. It explained the gaps and limitations and it mentioned the small sample and lack of controls and what they imply. I'd expect no less from Hannah Barnes - she wrote "Time to Think".

Though we can't stop other people "drawing inferences" to suit themselves.

Publishing null findings is good science. Publishing small-n studies with no control group then allowing people to draw inferences not supported by the data, not so much.

This is hugely important. The original paper with the original data was published a matter of days after the Bell vs Tavistock judgment, when the authors told the judge during the hearing it wasn’t ready yet. Then, even though the original data showed no improvement in psychological functioning, there were media organisations that claimed it showed patients were happier.

This data and its analysis represents all that is wrong with puberty blocking children: poor initial study format, no controls, poor outcomes, poor conclusions, and media spin on the outcomes. It’s an exemplar of every single paper purporting to show benefits of medically altering the bodies of children.

So the fact the data has been reanalysed to offer greater scrutiny is refreshing, and hopefully going to be also applied to every other paper. It happens with other papers but such analysis is either ignored, refused (Michael Biggs had an online comment removed analysing Jack Turban’s bullshit) or don’t reach wider attention because it’s published in a quiet space, like a lesser known journal.

This paper and its reporting on the BBC is sunlight.

Agreed. Given how fraught some of the discourse can be in both directions about transgender medicine and given the vagueness of this result, I'm inclined to be wary

Ah, yes - I'd mostly agree with you there.

There's still a place for 'X doesnt work' reports in the mainstream media if it's a reasonably clear result; less so for this type of thing. And the findings and significance need to be clearly explained, which this won't be.

I think you've misunderstood me - publishing no-result data is important, it's publicising it in the media and suggesting there is a result even where there doesn't seem to be is what I'm unsure of.

Publishing 'no result' data is massively important. Not doing so is one-of the big problems with research at the moment - only publishing results that show a strong effect skews the data.

I don't think it's that strange. Some saw no change, some an improvement, some saw a deterioration. So "change" is probably shorthand for "some improvements and some deterioration".

Regardless, I think it's irresponsible to have a headline which suggests a link to the drugs when the paper and indeed the article appended to said headline suggest that there's no known link at the moment between these drugs and mental health improving / worsening.

Its about as useful as saying "some people taking paracetamol went out and bought cars and some didn't".

I appreciate what this paper is trying to achieve, but as PP has said the source quality of the data is not great, and I'm really not sure how helpful publicising this is given that the results, even if they were obtained through rigourous controlled testing, appear fairly inconclusive with a roughly 1/3 breakdown between improvement, no change, and deterioration. Its all rather odd.

I thought the article title was odd

What a strange headline on the BBC article ("Children on puberty blockers saw mental health change").

If the trial had shown massive improvement in mental health, I'm sure they would be shouting that from the rooftops.

Thanks for the clarification.

I'm strongly in favour of accurate information and in-depth discussion on the use of these drugs, but this is misinformation. GnRH analogues were first developed in 1971, almost 20 years after Turing's death. He was given DES, a synthetic oestrogen and a different class of drug entirely.
There's a great deal wrong with GnRH analogues and how they're used - but that doesn't include high toxicity. They are simply not toxic in the way that many drugs, including paracetamol, are.

To be clear, I think there's strong evidence to support strictly limiting their use in children. But I feel people should be careful to avoid repeating sensationalist misinformation.

It looks an interesting paper. I'll give it a full read later, but I think the first important point to note is that this is not a new clinical trial. It's a re-analysis of a pre-existing data set from.somebody else's trial, to investigate whether using different analysis methods can extract better information from poor (uncontrolled and small sample size) trials.