This was posted on Twitter by Jolyon Maugham.
He has helpfully posted a link to the paper ( shop.bps.org.uk/publications/journals-and-periodicals/counselling-psychology-review-vol-34-no-1-june-2019.html ) but you need to pay to view it.
The abstract states that they randomly chose 303 patient files, followed them up and found one case of detransition. They conclude that as detransition is so low there is no need to slow the "pathway".
I'd be really interested to know how rigourous the methodology is - for example the abstract does not state how many of the random patients were actually tracked down.
I know some folk on here are really good at critically analysing and deconstruction journal articles....