Headline Based Medicine

Making waves with implausibly low levels of transition regret

One of the hottest political debates at the moment - especially in the US - is the cross-sex medicalisation of children. Even as the NHS in the UK starts to back away from this, publishing guidelines that it will no longer routinely prescribe puberty blockers, the US is split along partisan lines, with Republicans attempting to criminalise provision of youth gender medicine and Democrats attempting to criminalise obstructing it, or so it seems.

The possibility that a child might regret blockers and hormones and surgery in adulthood is something that is downplayed by proponents of paediatric transition. Time and again we hear that children know who they are, and that vanishingly few regret medically transitioning, and it is the only alternative to suicide. However there is precious little high quality data on regret, especially among this youth cohort being treated affirmatively.

So, into this fraught political arena steps a brand new study looking into rates of regret amongst women who have received “gender-affirmative” mastectomies over the last 30 years.

The conclusion states:

we observed long-term low levels of decisional regret and high satisfaction

And:

these results are consistent with previous ad hoc studies and affirm the overwhelmingly low levels of regret following gender-affirming surgery.

Sounds impressive - just how overwhelming were the results?

0

That’s right, this study seems to have found 0 regret.

The median decisional regret score was 0 on a 100-point scale, with lower scores indicating lower levels of regret.

Back of the net, another win for “gender affirmative” medicine, right? High satisfaction, no regrets, across the board. They didn’t just report no regret, these study participants absolutely aced it, with more than 75% of respondents giving the best possible scores for regret and satisfaction.

Alarm bells ought to be ringing at how implausible these findings are, but if you are a major media outlet reporting in the contested political space of transgender medical interventions, the best thing to do is to just go right ahead and reprint these claims as glowingly as possible:

The article contains neutral and responsible reporting like:

Some arguments in favor of laws that restrict gender-affirming care claim that patients may some day regret any irreversible or semi-irreversible part of their transition, but a small new study found that satisfaction with one such surgery is significant, even over the long term.

And:

A gender-affirming surgery can be transformative for an adult and may even be life-saving.

Yes, these mastectomies are life-saving, nobody regrets them, so attempts to ban medical interventions on minors are contradicted by science. Apparently.

So let’s look in a bit more detail as to why things might not be so joyous as the headlines make out.

Non Respondents

The first thing to note is that - common to so many studies on transition regret - this one suffers from a high rate of non-response, specifically: 41% of those contacted did not participate.

Now, it may be that all of those 41% are completely satisfied with their surgical outcomes - however, other studies have found that those with transition regret do not return to their healthcare providers, who they see as having facilitated harm.

The study itself was carried out by surgeons working in this area, and funded by The Plastic Surgery Foundation:

This work was supported by a pilot grant (923995) from The Plastic Surgery Foundation

…

The University of Michigan institutional review board requested that the authors also disclose their participation in the performance of gender-affirming surgery in their practice

None of this is necessarily a cause to disbelieve its results, but it does highlight a basic methodological weakness in that those conducting it are those who may have caused the regret in the first place, making it more likely that those with regret will not respond.

The study does publish a demographic breakdown between the respondents and non-respondents, and there seems to be very little difference between the two groups, with a notable exception: the more time had passed since surgery, the more likely a patient was to be non-responsive.

There are multiple possible explanations for this, but it cannot be ruled out that the more time has elapsed, the more likely someone is to start experiencing regret, and therefore the less likely someone is to respond to their surgeons when they come asking questions. Other studies have found that transition regret increases as a function of time, and that average time to regret is typically 5-7 years, after a period of initial euphoria.

It is wrong to assume that all 41% of the non-responders will regret their surgery, but it is also wrong to simply discount the possibility that regret may be highest in that group, and such regret can very commonly manifest as refusal to participate.

Median Regret

The regret numbers are presented in a strange way. Rather than highlighting how many individuals experienced regret, perhaps partitioned by severity, the study actually publishes only the median results, ie that the overwhelming majority experienced no regret.

The actual figures aren’t published, but if you look at the figure from the paper, something like 2 respondents experienced a level of regret above the halfway mark (one around 65/100 and the other around 90/100).

That’s about 1.4% of respondents reporting a significant amount of regret. Again, still small numbers, but more than the zero median figure.

This is significant because comparisons have to be drawn to other surgical procedures which all report rates of regret, not median regret levels.

Indeed, in an editorial comment published alongside the study, contributors affiliated with the Division of Plastic and Reconstructive Surgery at Rush University Medical Center, Chicago, said:

This study contradicts claims that regret following gender-affirming surgery may not manifest for many years and highlights the disproportionate criticism encountered in gender-affirming care compared with other surgical disciplines.

And also noted that, while most other surgeries have a regret rate of around 14%:

Interestingly, no legislative efforts are aimed at banning these procedures.

But since this study did not actually report the rate of regret (1.4%) and instead reported the 0 median it is inappropriate to draw such comparisons.

Excluded Regret

On top of this, of the 139 who did respond, 13 who indicated they did experience regret were excluded.

That’s over 9%.

So even if you ignore the 41% of non-respondents, almost one in ten of the participants did indicate regret but this was left out because the authors could not reconcile a high regret score with a high satisfaction score:

these 13 participants were noted to have high satisfaction and high regret, suggesting a possible error due to item order. This concern is further compounded by the established high correlation between satisfaction with decision and decisional regret. For this reason, these individuals were excluded from the analysis.

So because they expected those with high regret to also be dissatisfied with the outcome, anyone saying different was excluded as a mistake. Never mind all the other reasons this could be true (eg. someone experiencing regret but attempting to make peace with their new, permanent circumstances).

We don’t have many details on the ones who were excluded, but the paper notes that 4 of the 13 had a satisfaction score of 4.75 or higher alongside decision regret scores in the 95-100 range. To understand these scores, it is worth examining the different sets of questions.

The Decision Regret Scale is very clear and looks like this (with each question given a score, and the total summed, averaged and scaled to 100):

1. It was the right decision

2. I regret the choice that was made

3. I would go for the same choice if I had to do it over again

4. The choice did me a lot of harm

5. The decision was a wise one

It is quite easy to see how someone experiencing regret and harm would put maximum answers to all questions here, so four individuals scoring 95-100 on this scale is very clear regret.

However, satisfaction was measured with the Holmes-Rovner scale, which has six much more ambiguous questions (tailored for treatment), answered on a 1-5 scale (summed and averaged). The paper reports they only used four questions (though they don’t state which four) but one of their citations lists the following subset:

1. I was adequately informed about the issues

2. The decision about surgery was the best decision for me

3. The surgery decision was consistent with my personal values

4. I am satisfied with my decision about what type of surgery to have

We can’t know if it was exactly these questions, but it seems likely - and they are far more open to interpretation than the regret scale. Someone with regret that was unrelated to a lack of information or input beforehand, could very easily max out questions 1 and 3.

Question 2 is more ambiguous - if for example an individual is making peace with their present situation and rationalising it as the best possible choice they could have made at the time, they could understandably regret their present situation while believing they did their best under the circumstances. Anything from neutral to positive responses here are completely plausible.

Question 4 is the only one that is really straightforwardly in conflict with regret, but it is still possible to get 3 points on that question from a neutral answer, and again it could be open to interpretation (ie, perhaps they are relieved they didn’t go for an alternative surgical option, or are satisfied that they elected for a surgery that preserved nipple sensation).

On this scale, answering questions 1 and 3 positively, and 2 and 4 neutrally gives a score of 4.

If question 2 is answered positively then that score can be 4.5.

If nine respondents were excluded because their satisfaction scores were in that range, I think it is fair to say that “error” isn’t the only explanation for that incongruence and that it was wrong to exclude them. Even considering the four most incongruous scores - where satisfaction was 4.75 or higher - it seems wrong to discount someone scoring 95-100 on the far less ambiguous regret scale because of it.

There are plausible reasons why this is apparent disconnect is not necessarily an error, and it does point to the limited utility of these existing tools for analysing the complexity of feelings individuals can experience post-transition. Rather than note this apparent incongruence between regret and the satisfaction scores as part of the weakness of these standard tools, they simply assumed it must be an error and excluded the responses.

I think a good argument can be made that this was inappropriate and that - even disregarding the number of nonrespondents - when adding the unreported and excluded regret rates this study actually indicates a value somewhere between 8% and 10.5%.

Participant Age

Another significant limitation of the study is that it is solely on adults. The study recruited over-18s only, and most participants (both responders and non-responders) fell in the age range 23-33, with a median of around 26.

This is of course a vastly different cohort to minors and so any findings here won’t translate to that group. There is absolutely no reason to think that regret rates for cosmetic double mastectomies by 33-year-old women are going to be at all indicative of those for 14-year-old girls. We know that, given time, up to 90% of young people experiencing gender dysphoria have historically desisted after adolescence, with the majority simply growing up to be gay or lesbian adults.

This means that this study is only looking at that historic 10% who persisted into adulthood, and it cannot be used to draw any conclusions about likely regret rates in today’s children.

Since the overwhelming focus of attempts to ban such surgical interventions is focused on children, it would seem to make the CNN article’s placing this study in that context disingenuous. Yet can you really blame them when the study sets out in the abstract:

There has been increasing legislative interest in regulating gender-affirming surgery, in part due to the concern about decisional regret.

The authors clearly want this study to be seen as responsive to the hot political topic of transition regret in the context of legislative bans - which overwhelmingly affect minors - yet it sheds absolutely no light on paediatric interventions.

Not only that, the cohort of adults treated surgically between 1990 and 2020 is not remotely like the massive surge in young teens being referred for hormonal interventions under an informed consent/affirmative model. The therapeutic approach and level of gatekeeping has changed significantly in the last thirty years.

Multiple Genders

14% of those who responded had a stated gender identity other than “male” at their last medical encounter. That is, they were (as stated in the paper) non-binary, demiboy, genderqueer, multiple or some other gender.

However, this had risen to 39% by the time of responding to the survey. The study notes this included:

individuals reporting transgender and “nothing but confused at this point”:

The whole point of these procedures was to perform permanent surgical interventions on women who felt they were men - to the extent that they are recorded as “male” in their medical history - and yet their identity remains in a state of flux in subsequent years, even shifting to outright confusion. Is this a lifesaving mental health intervention or a lifestyle choice? Arguments against therapeutic gatekeeping and stricter regulation treat it as the only alternative to suicide, but massively uncertain outcomes like this are glibly considered successful.

I also note that 2 of the non-respondents reported their gender as “female” at the last medical contact. We don’t know what that contact was or when it took place, or how they identified prior, but it is strange for a woman to undergo surgery in pursuit of a “male gender identity” and yet still identify as a woman. Again, since they were nonresponsive, a complete binary detransition cannot be ruled out.

Multiple Surgeries

At least 9% of respondents (and 14% of non-respondents) had received at least one other surgery prior to receiving a mastectomy, mostly hysterectomies. One non-respondent had even undergone phalloplasty before mastectomy.

This is again, highly unusual when compared to the current cohort of young teenage girls, whose typical path is puberty blockers, hormones and mastectomy as first surgery.

It is also hard to tease out generalised regret over mastectomies specifically, when so many other procedures are in play, and 25% of respondents went on to further interventions afterwards. Since the timing of subsequent interventions is not recorded, this muddies the picture for establishing any relationship between regret and time since surgery.

Comorbid Conditions

One other area of difference between the respondents and non-respondents is mental health. Both sets had high rates of depression and anxiety at the time of surgery, but this was noticeably higher in the respondent group - around 70% compared to 44%.

There are several possible explanations for this, and it may well be an artefact of changing treatment pathways over the years, ie in years prior, surgical interventions were not permitted until underlying mental health issues were addressed.

In recent years, “gender affirmative medicine” has come to be seen in and of itself as a treatment for the depression and anxiety that arises from gender dysphoria. This would certainly explain why more recent cohorts tend to have worse mental health at the time of surgery.

Conclusion

This study is one that could have raised interesting questions, but instead races to a conclusion guaranteed to attract glowing headlines.

Rather than trumpeting a 0 median figure it could have published the up to 10.5% regret rate it actually found, and highlighted the apparent problems with the generalised satisfaction/regret tools that were used. The 41% non-response rate could have been used to highlight the possibility of using third-parties unconnected to any surgical or medical provider to conduct regret studies, to maximise the likelihood of obtaining a response from those with deep regret who wanted no further contact from their plastic surgeons. The rising comorbidities in those who have been treated more recently could have highlighted the recent changing approach to reduced therapeutic gatekeeping before surgery. Clear distinctions between this adult cohort and the ethical minefield of performing permanent surgical interventions on children could have been made, instead of heavily implying an equivalence. The increasing fluidity of identity of participants - and in some cases outright confusion - could all have brought into question why these interventions are even happening in the first place.

Most importantly it could have approached the issue of the regret and grief of those who feel this was a path they never needed to go down with some nuance and compassion. Instead it seems like every possible move has been made to downplay the possibility of regret.

There is useful information here that could point to areas of future study, but most seems to be disregarded in an attempt to paint an unrealistic picture. Can CNN really be blamed for taking such a rosy view and tying this study into the wider political landscape of bans on youth gender medicine when the study itself positions itself that way? Indeed, in the accompanying editorial comment the implausible results are presented in overconfident fashion, and concerns are downplayed and dismissed as illegitimate. Criticism is “disproportionate”.

It is clear that this adult study is going to become simply more ammunition in a political battle over regulation of youth medical transition. Few will read past the headline “no regrets”, and anyone who does will be condemned as a bigot or right-wing or of fearmongering.

And meanwhile, those actually experiencing transition grief are appropriated or ignored in equal measure, dismissed or exploited according to whichever way the political wind blows. Rendered invisible, reimagined as steps on a “transition journey”, or described in dehumanising terms by those who appropriate their experiences, their surgeries, their bodies as shock tactics in pursuit of moral outrage and political pointscoring.

I think they deserve better than this shabby mess.