Letter to the Editor: Regret after Gender-affirmation Surgery: A Systematic Review and Meta-analysis of Prevalence.

Sir,

Bustos et al[1] aimed to measure the prevalence of regret following gender-affirmation surgery. Given the significant rise in young people seeking medical intervention for gender dysphoria, which can include surgery, outcome studies that accurately assess regret are of increasing importance. In this letter, we argue that the conclusions of their systematic review and meta-analysis are questionable due to limitations in their methods and shortcomings of the studies selected.

Starting with methods, the authors overlooked numerous relevant studies, including one of the best-known,[2] raising questions about the adequacy of their search strategy. One study[3] was inappropriately included as it only investigated regret regarding choice of surgical procedure, not of surgery itself. In addition, there are significant data extraction errors, leading to erroneous conclusions. For instance, the sample for surgical regret in their largest included study[4] was inflated from 2627 to 4863, likely due to a miscalculation from a table reporting the treatment patterns of that paper’s total study population.

Besides these methodological inaccuracies, data in this field are often of low quality because of “lack of controlled studies, incomplete follow-up, and lack of valid assessment measures,”[5] as well as the long amount of time regret can take to manifest (the average and median are estimated at 8–8.5 years[2,4]). Many of the included studies had participants with follow-up periods of only 1 or 2 years postsurgical transition. None appear to have a long enough follow-up period to reliably identify regret. The study contributing almost half of the participants[4] explicitly noted their inclusion of participants with short follow-up time, relative to time to regret, and their large 36% loss to follow-up as limitations. These shorter studies only provide an estimated lower limit, as the large numbers of patients lost to follow-up add correspondingly large uncertainties to any quoted number.

Bustos et al[1] acknowledge “moderate-to-high risk of bias in some studies.” Actually, this affects 23 of the 27 studies. The majority of included studies ranged between “poor” and “fair” quality: only five studies—representing just 3% (174) of total participants—received higher quality ratings. However, even these had loss to follow-up rates ranging from 28% to more than 40%, including loss through death from complications or suicide, negative outcomes potentially associated with regret.

A last and major concern involves sample selection. The cohort presenting with gender dysphoria today is substantially different from the cohort presenting during the research periods of the included studies. Further, there has been a significant liberalization over time of the criteria assessing readiness for surgery. Thus, the outcomes reported may be of limited relevance for estimating current surgery outcomes. Additionally, the generalization to “TGNB” populations seems unreliable, as it is based on an explicit sample size of only one “non-binary” patient. The authors do not address these issues.

In light of these numerous issues affecting study quality and data analysis, their conclusion that “our study has shown a very low percentage of regret in TGNB population after GAS” is, in our opinion, unsupported and potentially inaccurate.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.