Small study size is a real issue and will continue to be
The original study was uncontrolled and has a small sample size. The very small sample size at 36 months is particularly important to bear in mind when interpreting results. The reason for the reduction in sample size is given as “people were recruited at different ages (12–15 years) but left the study soon after their 16th birthday”; no further explanation is given about why this was the case (Carmichael et al., Citation
2021a). While the individual level change approach can be applied irrespective of sample size and lack of a controlled design, there remains a fundamental problem that there is insufficient data to draw firm conclusions about the safety and efficacy of this treatment.
Relatively few participants fell into the borderline or clinical range at baseline (27%–58% depending on the scale). This is lower than a typical CAMHS population; Gibbons et al. (Citation
2021) found around 90% of their sample to be in the clinical or borderline range at baseline. This suggests that the GIDS participants may have been less distressed when they were referred to the clinic than typical CAMHS referrals and may indicate a relatively low referral threshold operating at the time period studied. I find this significant, referrals were clearly 'acceptance' at all costs
Here we have it, biological sex might be important after all. Who'd a thought
For example, the approach could be used to look more closely at biological sex in order to say, for example, that "x% of biological boys improved"; "x% of biological girls deteriorated"; “x% of biological boys experienced side effect A” and so on, which would improve consent processes in clinical practice until better data is available
Common sense has entered the room
For example, it would be possible to analyze routinely collected outcome data for children who do not take up puberty blockers to generate comparable data on Reliable and Clinically Significant Change as a naturalistic control group.