Thursday, 5 October 2017

Vanishingly small: "regret" statistics interrogated.

In a world dominated by post-truth politics even apparent hard figures can be misleading, and, for a general public that is relatively ignorant about statistics, it is easy to mislead while not technically lying. A headline like "Football violence doubles in three months!" may sound alarming but if that is an increase of 0.0001% to 0.0002% of supporters then it is very different from an increase from 5% to 10%. The headline would be true in both cases yet we would be looking at an epidemic in the second scenario but the statistical effect of possibly only one incident in the first.

Likewise with statistics about trans "regret".

The number of trans “regretters” is vanishingly small and difficult to ascertain as a percentage, so any kind of statistics about them are automatically going to be problematic. Statistics do not work with very small numbers, that is why opinion polls take a four-figure sample, and then they still usually get it wrong. So forensically examining any figures is what any responsible journalist should do before publication. The problem is that, in the current heavily biased anti-trans media onslaught, they are not doing this and consequently anything that is biased against trans people is automatically treated as true.

So the material produced in the media about a surgeon doing more trans reversals needs to be examined carefully. When we look at the figures provided in the press we can see the following;
Over 5 years he has had 13 trans people contact him about so-called reversal surgery (two are mentioned as being “in surgery” but it is not clear as to whether they are additional to the existing 13 or included in that number). That is approximately three a year. The first six are described as coming from “all over the “Western World” The “western World” presumably including most of Europe, north America and Australasia. For the other 7 it is not clear where they come from. One of the claims made in the article was that the "average" age of his clients had come down to 21. I find this to be a very spurious use of statistics. When you are talking about 13 people over 5 years, then just getting two 21-year-olds would be enough to change the "average". Of course we are not told the timescale for this "average" but if he gets two this year, or even just one, that could produce an "average" for this year of 21. we have also not been told whether it is a mean, mode or median average too, which could skew the stats. This is what I mean that using stats when the numbers you are talking about are tiny is totally meaningless and open to abuse, without technically lying.

Over the last year in the US, there have been 3,250 gender reassignment surgeries, in the UK there were 172 operations in 2014, with 280 on the list for surgery in 2017. Numbers are difficult to ascertain for other countries but we are probably looking at a higher figure than the 3,250 in the US for 2016 in Europe so we are probably looking at a figure of around 7,000 a year for the “Western World”. If we go beyond that to the Middle East, South Asia, the Asia-Pacific countries, Africa and Latin America, where the populations are larger, but whose access to surgery is restricted by poverty, the number of surgeries is probably running at well over 20,000 a year globally, it is probably difficult to tell exactly, but these are almost certainly conservative figures. Now there have been lots of stories about transgender surgery increasing exponentially, at least in the “Western World” so let us assume that the number was around 10,000 five years ago and has increased by around 2,000 a year in the intervening period. That means at least 80,000 people globally have had gender affirmation surgery in the last five years.

If we also remember that at least some of the patients in this surgeon’s figures must have had surgery more than 5 years ago then his clinic is probably dealing with a subset of patients from a group that exceeds 100,000 people. If this is the case then we are looking at 1 regretter for every 7692 people having gender affirmation surgery. In other words for every 7691 successful gender affirmation surgeries there is one unsuccessful one. This would give us a regret rate of less than 0.013%.

Now obviously I have to heed my own warning about statistics, so let us assume, for the sake of argument, that the global number of surgeries is half what I have estimated for the period in question; 50,000, that would mean that there is one regretter for every 3845 successful surgeries, still a vanishingly small regret rate. Even if we halve it again the regret rate comes out at 1922 successful surgeries for every regretter, a figure significantly less than 0.1%. Even if we then double the number of regretters there are still 960 successful operations for every regretter. Double it again and the ratio of successful operations to unsuccessful ones is still 480:1 a vanishingly small percentage. Let us remember that this figure is achieved by quadrupling the figures for this doctor and reducing the estimate for GRS by 75%, even doing this gives us a tiny percentage, well under 1%.

Of course one of the statistics we do not know is how many of those 100,000 successful operations would have resulted in death by suicide if they had been denied access to surgery. This is a statistic we can never obtain ethically of course. What is clear from those who are responsible for the gatekeeping processes for surgery is the way they consistently refer to the risk of suicide for their patients. The death of 15-year-old Leo Hetherington earlier this year, he after was told he could not have GRS, and from the people I know, especially young people, suggests that this figure would be very high indeed.

So media articles that talk about a rise in the numbers of “regretters” are effectively being dishonest and transphobic because they are not contextualising these figures. There is obviously going to be a rise in regretters because there is an increase in the number of surgeries. What matters is the proportion of regretters to successful surgeries and here it would appear that, if anything the percentage of them compared to the total number of surgeries performed, is getting significantly smaller. Without contextualising these figures they become meaningless at best and profoundly dishonest at worst. What journalists also need to do to contextualise these figures is to look at regret rates for other types of surgery. The lowest regret rate I could find was for LASIK eye surgery, which is described as having a “very low” dissatisfaction rate of 4%. Many other forms of surgery have dissatisfaction rates that are well into double figures. One of the types of surgery that many trans-haters try to liken to gender affirmation surgery, in order to delegitimise us, is cosmetic surgery. However the difference between the two is a whopping 65% regret rate for cosmetic surgery compared to a tiny fraction of 1% for transgender surgery.

The truth is that surgery for trans people is one of - if not the most - successful surgeries that is carried out on this planet with a regret rate that makes pretty much anything else look irresponsible and which saves countless lives. Some of those lives saved at the time will probably also include people who subsequently become regretters. As psychologist Diane Ehrensaft put it when talking about the risk of regret;

"Is it a risk? Let us call it a possibility. If that is a possibility we think that the most important thing is the same exact idea, find out who you are and make sure you get help facilitating being that person then. We have one risk we know about; the risk to youth when you hold them back and hold back those interventions; depression, anxiety, suicide attempts - even successes. And, if we can give them a better life I weigh that against that there may be a possibility of a change later, but they won’t be alive to change. That’s how I weigh up the scales."

No comments:

Post a Comment