Uhtred son of Uhtred, regular ale drinker, who I predict will die of injury (but will go to Valhalla, unlike you you ale-sodden wretch)

There has been some fuss in the media recently about a new study showing no level of alcohol use is safe. It received a lot of media attention (for example here), reversed a generally held belief that moderate consumption of alcohol improves health (this is even enshrined in the Greek food pyramid, which has a separate category for wine and olive oil[1]), and led to angsty editorials about “what is to be done” about alcohol. Although there are definitely things that need to be done about alcohol, prohibition is an incredibly stupid and dangerous policy, and so are some of its less odious cousins, so before we go full Leroy Jenkins on alcohol policy it might be a good idea to ask if this study is really the bees knees, and does it really show what it says it does.

This study is a product of the Global Burden of Disease (GBD) project, at the Institute for Health Metrics and Evaluation (IHME). I’m intimately acquainted with this group because I made the mistake of getting involved with them a few years ago (I’m not now) so I saw how their sausage is made, and I learnt about a few of their key techniques. In fact I supervised a student who, to the best of my knowledge, remains the only person on earth (i.e. the only person in a population of 7 billion people, outside of two people at IHME) who was able to install a fundamental software package they use. So I think I know something about how this institution does its analyses. I think it’s safe to say that they aren’t all they’re cracked up to be, and I want to explain in this post how their paper is a disaster for public health.

The way that the IHME works in these papers is always pretty similar, and this paper is no exception. First they identify a set of diseases and health conditions related to their chosen risk (in this case the chosen risk is alcohol). Then they run through a bunch of previously published studies to identify the numerical magnitude of increased risk of these diseases associated with exposure to the risk. Then they estimate the level of exposure in every country on earth (this is a very difficult task which they use dodgy methods to complete). Then they calculate the number of deaths due to the conditions associated with this risk (this is also an incredibly difficult task to which they apply a set of poorly-accredited methods). Finally they use a method called comparative risk assessment (CRA) to calculate the proportion of deaths due to the exposure. CRA is in principle an excellent technique but there are certain aspects of their application of it that are particularly shonky, but which we probably don’t need to touch on here.

So in assessing this paper we need to consider three main issues: how they assess risk, how they assess exposure, and how they assess deaths. We will look at these three parts of their method and see that they are fundamentally flawed.

Problems with risk assessment

To assess the risk associated with alcohol consumption the IHME used a standard technique called meta-analysis. In essence a meta-analysis collects all the studies that relate an exposure (such as alcohol consumption) to an outcome (any health condition, but death is common), and then combines them to obtain a single final estimate of what the numerical risk is. Typically a meta-analysis will weight all the risks from all the studies according to the sample size of the study, so that for example a small study that finds banging your head on a wall reduces your risk of brain damage is given less weight in the meta-analysis than a very large study of banging your head on a wall. Meta-analysis isn’t easy for a lot of reasons to do with the practical details of studies (for example if two groups study banging your head on a wall do they use the same definition of brain damage and the same definition of banging?), but once you iron out all the issues it’s the only method we have for coming to comprehensive decisions about all the studies available. It’s important because the research literature on any issue typically includes a bunch of small shitty studies, and a few high quality studies, and we need to balance them all out when we assess the outcome. As an example, consider football and concussion. A good study would follow NFL players for several seasons, taking into account their position, the number of games they played, and the team they were in, and compare them against a concussion free sport like tennis, but matching them to players of similar age, race, socioeconomic background etc. Many studies might not do this – for example a study might take 20 NFL players who died of brain injuries and compare them with 40 non-NFL players who died of a heart attack. A good meta-analysis handles these issues of quality and combines multiple studies together to calculate a final estimate of risk.

The IHME study provides a meta-analysis of all the relationships between alcohol consumption and disease outcomes, described as follows[2]:

we performed a systematic review of literature published between January 1st, 1950 and Dec 31st 2016 using Pubmed and the GHDx. Studies were included if the following conditions were met. Studies were excluded if any of the following conditions were met:

1. The study did not report on the association between alcohol use and one of the included outcomes.

2. The study design was not either a cohort, case-control, or case-crossover.

3. The study did not report a relative measure of risk (either relative risk, risk ratio, odds-ratio, or hazard ratio) and did not report cases and non-cases among those exposed and un-exposed.

4. The study did not report dose-response amounts on alcohol use.

5. The study endpoint did not meet the case definition used in GBD 2016.

There are many, many problems with this description of the meta-analysis. First of all they seem not to have described the inclusion criteria (they say “Studies were included if the following conditions were met” but don’t say what those conditions were). But more importantly their conditions for exclusion are very weak. We do not, usually, include case-control and case-crossover studies in a meta-analysis because these studies are, frankly, terrible. The standard method for including a study in a meta-analysis is to assess it according to the Risk of Bias Tool and dump it if it is highly biased. For example, should we include a study that is not a randomized controlled trial? Should we include studies where subjects know their assignment? The meta-analysis community have developed a set of tools for deciding which studies to include, and the IHME crew haven’t used them.

This got me thinking that perhaps the IHME crew have been, shall we say, a little sloppy in how they include studies, so I had a bit of a look. On page 53-55 of the appendix they report the results of their meta-analysis of the relationship between atrial fibrillation and alcohol consumption, and the results are telling. They found 9 studies to include in their meta-analysis but there are many problems with these studies. One (Cohen 1988) is a cross-sectional study and should not be included, according to the IHME’s own exclusion criteria. 6 of the remaining studies assess fribillation only, while 2 assess fibrillation and fibrial flutter, a pre-cursor of fibrillation. However most tellingly, all of these studies find no relationship between alcohol consumption and fibrillation at almost all levels of consumption, but their chart on page 54 shows that their meta-analysis found an almost exponential relationship between alcohol consumption and fibrillation. This finding is simply impossible given the observed studies. All 9 studies found no relationship between moderate alcohol consumption and fibrillation, and several found no relationship even for extreme levels of consumption, but somehow the IHME found a clear relationship. How is this possible?

Problems with exposure assessment

This problem happened because they applied a tool called DISMOD to the data to estimate the relationship between alcohol exposure and fibrillation. DISMOD is an interesting tool but it has many flaws. Its main benefit is that it enables the user to incorporate exposures that have many different categories of exposure definition that don’t match, and turn them into a single risk curve. So for example if one study group has recorded the relative risk of death for 2-5 drinks, and another group has recorded the risk for 1-12 drinks, DISMOD offers a method to turn this into a single curve that will represent the risk relationship per additional drink. This is nice, and it produces the curve on page 54 (and all the subsequent curves). It’s also bullshit. I have worked with DISMOD and it has many, many problems. It is incomprehensible to everyone except the two guys who programmed it, who are nice guys but can’t give decent support or explanations of what it does. It has a very strange response distribution and doesn’t appear to apply other distributions well, and it has some really kooky Bayesian applications built in. It is also completely inscrutable to 99.99% of people who use it, including the people at IHME. It should not be used until it is peer reviewed and exposed to a proper independent assessment. It is application of DISMOD to data that obviously shows no relationship between alcohol consumption and fibrillation that led to the bullshit curve on page 54 of the appendix, that does not have any relationship to the observed data in the collected studies.

This also applies to the assessment of exposure to alcohol. The study used DISMOD to calculate each country’s level of individual alcohol consumption, which means that the same dodgy technique was applied to national alcohol consumption data. But let’s not get hung up on DISMOD. What data were they using? The maps in the Lancet paper show estimates of risk for every African and south east Asian country, which suggests that they have data on these countries, but do you think they do? Do you think Niger has accurate estimates of alcohol consumption in its borders? No, it doesn’t. A few countries in Africa do and the IHME crew used some spatial smoothing techniques (never clearly explained) to estimate the consumption rates in other countries. This is a massive dodge that the IHME apply, which they call “borrowing strength.” At its most egregious this is close to simply inventing data – in an earlier paper (perhaps in 2012) they were able to estimate rates of depression and depression-related conditions for 183 (I think) countries using data from 97 countries. No prizes to you, my astute reader, if you guess that all the missing data was in Africa. The same applies to the risk exposure estimates in this paper – they’re a complete fiction. Sure for the UK and Australia, where alcohol is basically a controlled drug, they are super accurate. But in the rest of the world, not so much.

Problems with mortality assessment

The IHME has a particularly nasty and tricky method for calculating the burden of disease, based around a thing called the year of life lost (YLL). Basically instead of measuring deaths they measure the years of your life that you lost when you died, compared to an objective global standard of life you could achieve. Basically they get the age you died, subtract it from the life expectancy of an Icelandic or Japanese woman, and that’s the number of YLLs you suffered. Add that up for every death and you have your burden of disease. It’s a nice idea except that there are two huge problems:

  • It weights death at young ages massively
  • They never incorporate uncertainty in the ideal life expectancy of an Icelandic or Japanese woman

There is an additional problem in the assessment of mortality, which the IHME crew always gloss over, which is called “garbage code redistribution.” Basically, about 30% of every country’s death records are bullshit, and don’t correspond with any meaningful cause of death. The IHME has a complicated, proprietary system that they cannot and will not explain that redistributes these garbage codes into other meaningful categories. What they should do is treat these redistributed deaths as a source of error (e.g. we have 100,000 deaths due to cancer and 5,000 redistributed deaths, so we actually have 102500 plus/minus 2500 deaths), but they don’t, they just add them on. So when they calculate burden of disease they use the following four steps:

  • Calculate the raw number of deaths, with an estimate of error
  • Reassign dodgy deaths in an arbitrary way, without counting these deaths as any form of uncertainty
  • Estimate an ideal life expectancy without applying any measure of error or uncertainty to it
  • Calculate the years of life lost relative to this ideal life expectancy and add them up

So here there are three sources of uncertainty (deaths, redistribution, ideal life expectancy) and only one is counted; and then all these uncertain deaths are multiplied by the number of years lost relative to the ideal life expectancy.

The result is a dog’s breakfast of mortality estimates, that don’t come even close to representing the truth about the burden of disease in any country due to any condition.

Also, the IHME apply the same dodgy modeling methods to deaths (using a method that they (used to?) call CoDMoD) before they calculate YLLs, so there’s another form of arbitrary model decisions and error in their assessments.

Putting all these errors together

This means that the IHME process works like this:

  • An incredibly dodgy form of meta-analysis that includes dodgy studies and miscalculates levels of risk
  • Applied to a really shonky estimate of the level of exposure to alcohol, that uses a computer program no one understands applied to a substandard data set
  • Applied to a dodgy death model that doesn’t include a lot of measures of uncertainty, and is thus spuriously accurate

The result is that at every stage of the process the IHME is unreasonably confident about the quality of their estimates, produces excessive estimates of risk and inaccurate measures of exposure, and is too precise in its calculations of how many people died. This means that all their conclusions about the actual risk of alcohol, the level of exposure, and the magnitude of disease burden due to the conditions they describe cannot be trusted. As a result, neither can their estimates of the proportion of mortality due to alcohol.

Conclusion

There is still no evidence that moderate alcohol consumption is bad for you, and solid meta-analyses of available studies support the conclusion that moderate alcohol consumption is not harmful. This study should not be believed and although the IHME has good press contacts, you should ignore all the media on this. As a former insider in the GBD process I can also suggest that in future you ignore all work from the Global Burden of Disease project. They have a preferential publishing deal with the Lancet, which means they aren’t properly peer reviewed, and their work is so massive that it’s hard for most academics to provide adequate peer review. Their methods haven’t been subjected to proper external assessment and my judgement, based on having visited them and worked with their statisticians and their software, is that their methods are not assessable. Their data is certainly dubious at times but most importantly their analysis approach is not correct and the Lancet doesn’t subject it to proper peer review. This is going to have long term consequences for global health, and at some point the people who continue to associate with the IHME’s papers (they have hundreds or even thousands of co-authors) will regret that association. I stopped collaborating with this project, and so should you. If you aren’t sure why, this paper on alcohol is a good example.

So chill, have another drink, and worry about whether it’s making you fat.


fn1: There are no reasons not to love Greek food, no wonder these people conquered the Mediterranean and developed philosophy and democracy!

fn2: This is in the appendix to their study

Advertisements

No this really is not “the healthy one”

Today’s Guardian has a column by George Monbiot discussing the issue of obesity in modern England, that I think fundamentally misunderstands the causes of obesity and paints a dangerously rosy picture of Britain’s dietary situation. The column was spurred by a picture of a Brighton Beach in 1976, in which everyone was thin, and a subsequent debate on social media about the causes of the changes in British rates of overweight and obesity in the succeeding half a decade. Monbiot’s column dismisses the possibility that the growth in obesity could be caused by an increase in the amount we eat, by a reduction in the amount of physical activity, or by a change in rates of manual labour. He seems to finish the column by suggesting it is all the food industry’s fault, but having dismissed the idea that the food industry has convinced us to eat more, he is left with the idea that the real cause of obesity is changes in the patterns of what we eat – from complex carbohydrates and proteins to sugar. This is a bugbear of certain anti-obesity campaigners, and it’s wrong, as is the idea that obesity is all about willpower, which Monbiot also attacks. The problem here though is that Monbiot misunderstands the statistics badly, and as a result dismisses the obvious possibility that British people eat too much. He commits two mistakes in his article: first he misunderstands the statistics on British food consumption, and secondly he misunderstands the difference between a rate and a budget, which is ironic given he understands these things perfectly well when he comments on global warming. Let’s consider each of these issues in turn.

Misreading the statistics

Admirably, Monbiot digs up some stats from 1976 and compares them with statistics from 2018, and comments:

So here’s the first big surprise: we ate more in 1976. According to government figures, we currently consume an average of 2,130 kilocalories a day, a figure that appears to include sweets and alcohol. But in 1976, we consumed 2,280 kcal excluding alcohol and sweets, or 2,590 kcal when they’re included. I have found no reason to disbelieve the figures.

This is wrong. Using the 1976 data, Monbiot appears to be referring to Table 20 on page 77, which indicates a yearly average of 2280 kCal. But this is the average per household member, and does not account for whether or not a household member is a child. If we refer to Table 24 on page 87, we find that a single adult in 1976 ate an average of 2670 kCal; similar figures apply for two adult households with no children (2610 kCal). Using the more recent data Monbiot links to, we can see that he got his 2,130 kCal from the file of “Household and Eating Out Nutrient Intakes”. But if we use the file “HC – Household nutrient intakes” and look at 2016/17 for households with one adult and no children, we find 2291 kCal, and about 2400 as recently as 10 years ago. These are large differences when they accrue over years.

This is further compounded by the age issue. When we look at individual intake we need to consider how old the family members are. If an average individual intake is 2590 kCal in 1976 including alcohol and sweets, as Monbiot suggests, we need to rebalance it for adults and children. In a household with three people we have 7700 kCal, which if the child is eating 1500 kCal means that the adults are eating close to 3100 kCal each. That’s too much food for everyone in the house, even using the ridiculously excessive nutrient standards provided by the ONS.  It’s also worth remembering that the age of adults in 1976 was on average much younger than now, and an intake of 2590 might be okay for a young adult but it’s not okay for a 40-plus adult, of which there are many more now than there were then. This affects obesity statistics.

Finally it’s also worth remembering that obesity is not evenly distributed, and an average intake of 2100 kCal could correspond to an average of 2500 in the poorest 20% of the population (where obesity is common) and 1700 kCal in the richest, which is older and thinner. An evenly distributed 2100 kCal will lead to zero obesity over the whole population, but an unevenly distributed 2100 kCal will not. It’s important to look carefully at the variation in the datasets before deciding the average is okay.

Misunderstanding budgets and rates

Let’s consider the 2590 kCal that Monbiot finds as the average intake of adults in 1976, including alcohol and sweets. This is likely wrong, and the average is probably more like 3000 kCal including alcohol and sweets, but let’s go with it for now. Monbiot is looking to see what has changed in our diet over the past 40 years to lead to current rates of obesity, because he is looking for a change in the rate of consumption. But he doesn’t consider that all humans have a budget, and that a small excess of that budget over a long period is what drives obesity. The reality is that today’s obesity rates do not reflect today’s consumption rates, but the steady pattern of consumption over the past 40 years. What made a 55 year old obese today is what they ate in 1976 – when they were 15 – not what the average person eats today. So rather than saying “we eat less today than we did 40 years ago so that can’t be the cause of obesity”, what really matters is what people have been eating for the past 40 years. And the stats Monbiot uses suggest that women, at least, have been eating too much – a healthy adult woman should eat about 2100 kCal, and if the average is 2590 then a woman in 1976 has been at or above her energy intake every year for the past 40 years. It doesn’t matter that a woman’s intake declined to 2100 kCal in 2016, because she has been eating too much for the past 35 years anyway. It’s this budget, not changes over time, which determine the obesity rate now, and Monbiot is wrong to argue that it’s not overeating that has caused the obesity epidemic. Unless he accepts that a woman can eat 2590 kCal every year for 40 years and stay thin, he needs to accept that the problem of obesity is one of British food culture over half a century.

What this means for obesity policy

Somewhat disappointingly and unusually for a Monbiot article, there are no sensible policy prescriptions at the end except “stop shaming fat people.” This isn’t very helpful, and neither is it helpful to dismiss overeating as a cause, since everyone in public health knows that overeating is the cause of obesity. For example, Public Health for England wants to reduce British calorie intake, and the figures on why are disturbing reading. Reducing calorie intake doesn’t require shaming fat people but it does require acknowledgement that British people eat too much. This comes down not to individual willpower but to the food environment in which we all make choices about what to eat. The simplest way, for example, to reduce the amount that people eat is not to give them too much food. But there is simply no way in Britain that you can eat out or buy packaged food products without buying too much food. It is patently obvious that British restaurants serve too much food, that British supermarkets sell food in packages that are too large, and that as a result the only way for British people not to eat too much is through constant acts of will – leaving half the food you paid for, buying only fresh food in small amounts every day (which is only possible in certain wealthy inner city suburbs), and carefully controlling where, when and how you eat. This is possible but it requires either that you move in a very wealthy cultural circle where the environment supports this kind of thing, or that you personally exert constant control over your life. And that latter choice will inevitably end in failure, because constantly controlling every aspect of your food intake in opposition to the environment where you purchase, prepare and consume food is very very difficult.

When you live in Japan you live in a different food environment, which encourages small serving sizes, fresh and raw foods, and low fat and low sugar foods. In Japan you live in a food environment where you are always close to a small local supermarket with convenient opening hours and fresh foods, and where convenience stores sell healthy food in small serving sizes. This means that you can choose to buy small amounts of fresh food as and when you need them, and avoid buying in bulk in a pattern that encourages over consumption. When your food choices fail (for example you have to eat out, or buy junk food) you will have access to a small, healthy serving. If you are a woman you will likely have access to a “woman’s size” or “princess size” that means you can eat the smaller calorific food that your smaller calorific requirements suggest is wisest. It is easy to be thin in Japan, and so most people are thin. Overeating in Japan really genuinely is a choice that you have to choose to make, rather than the default setting. This difference in food environment is simple, obvious and especially noticeable when (as I just did) you hop on a plane to the UK and suddenly find yourself confronted with double helpings of everything, and super markets where everything is “family sized”. The change of food environment forces you to eat more. It’s as simple as that.

What Britain needs is a change in the food environment. And achieving a change in food environment requires first of all recognizing that British people eat too much, and have been eating too much for way too long. Monbiot’s article is an exercise in denialism of that simple fact, and he should change it or retract it.

The journal Molecular Autism this week published an article about the links between Hans Asperger and the Nazis in world war 2 Vienna, Austria. Hans Asperger is the paediatric pscyhiatrist on whose work Asperger’s syndrome is based, and after whom the syndrome is known. Until recently Asperger was believed to have been an anti-Nazi, someone who resisted the Nazis and risked his own career to protect some of his developmentally delayed patients from the Nazi “euthanasia” program, which killed or sterilized people with certain developmental disabilities for eugenics reasons.

The article, entitled Hans Asperger, National Socialism, and “race hygiene” in Nazi-era Vienna, is a thorough, well-researched and extensively documented piece of work, which I think is based on several years of detailed examination of primary sources, often in their original German. It uses these sources – often previously untouched – to explore and rebut several claims Asperger made about himself, and also to examine the nature of his diagnostic work during the Nazi era to see whether he was resisting or aiding the Nazis in their racial hygiene goals. In this post I want to talk a little about the background of the paper, and ask a few questions about the implications of these findings for our understanding of autism, and also for our practice as public health workers in the modern era. I want to make clear that I do not know much if anything about Asperger’s syndrome or autism, so my questions are questions, not statements of opinion disguised as questions.

What was known about Asperger

Most of Asperger’s history under the Nazis was not known in the English language press, and when his name was attached to the condition of Asperger’s syndrome he was presented as a valiant defender of his patients against Nazi racial hygiene, and as a conscientious objector to Nazi ideology. This view of his life was based on some speeches and written articles translated into English during the post war years, in particular a 1974 interview in which he claims to have defended his patients and had to be saved from being arrested by the Gestapo twice by his boss, Dr. Hamburger. Although some German language publications were more critical, in general Asperger’s statements about his own life’s work were taken at face value, and seminal works in 1981 and 1991 that introduced him to the medical fraternity did not include any particular reference to his activities in the Nazi era.

What Asperger actually did

Investigation of the original documents shows a different picture, however. Before Anschluss (the German occupation of Austria in 1938), Asperger was a member of several far right Catholic political organizations that were known to be anti-semitic and anti-democratic. After Anschluss he joined several Nazi organizations affiliated with the Nazi party. His boss at the clinic where he worked was Dr. Hamburger, who he claimed saved him twice from the Gestapo. In fact Hamburger was an avowed neo-nazi, probably an entryist to these Catholic social movements during the period when Nazism was outlawed in Vienna, and a virulent anti-semite. He drove Jews out of the clinic even before Anschluss, and after 1938 all Jews were purged from the clinic, leaving openings that enabled Asperger to get promoted. It is almost impossible given the power structures at the time that Asperger could have been promoted if he disagreed strongly with Hamburger’s politics, but we have more than circumstantial evidence that they agreed: the author of the article, Herwig Czech, uncovered the annual political reports submitted concerning Asperger by the Gestapo, and they consistently agreed that he was either neutral or positive towards Nazism. Over time these reports became more positive and confident. Also during the war era Asperger gained new roles in organizations outside his clinic, taking on greater responsibility for public health in Vienna, which would have been impossible if he were politically suspect, and his 1944 PhD thesis was approved by the Nazis.

A review of Asperger’s notes also finds that he did send at least some of his patients to the “euthanasia” program, and in at least one case records a conversation with a parent in which the child’s fate is pretty much accepted by both of them. The head of the institution that did the “euthanasia” killings was a former colleague of Asperger’s, and the author presents pretty damning evidence that Asperger must have known what would happen to the children he referred to the clinic. It is clear from his speeches and writings in the Nazi era that Asperger was not a rabid killer of children with developmental disabilities: he believed in rehabilitating children and finding ways to make them productive members of society, only sending the most “ineducable” children to institutional care and not always to the institution that killed them. But it is also clear that he accepted the importance of “euthanasia” in some instances. In one particularly compelling situation, he was put in charge – along with a group of his peers – of deciding the fate of some 200 “ineducable” children in an institution for the severely mentally disabled, and 35 of those ended up being murdered. It seems unlikely that he did not participate in this process.

The author also notes that in some cases Asperger’s prognoses for some children were more severe than those of the doctors at the institute that ran the “euthanasia” program, suggesting that he wasn’t just a fairweather friend of these racial hygiene ideals, and the author also makes the point that because Asperger remained in charge of the clinic in the post-war years he was in a very good position to sanitize his case notes of any connection with Nazis and especially with the murder of Jews. Certainly, the author does not credit Asperger’s claims that he was saved from the Gestapo by Hamburger, and suggests that these are straight-up fabrications intended to sanitize Asperger’s role in the wartime public health field.

Was Asperger’s treatment and research ethical in any way?

Reading the article, one question that occurred to me immediately was whether any of his treatments could be ethical, given the context, and also whether his research could possibly have been unbiased. The “euthanasia” program was actually well known in Austria at the time – so well known in fact that at one point allied bombers dropped leaflets about it on the town, and there were demonstrations against it at public buildings. So put yourself in the shoes of a parent of a child with a developmental disability, bringing your child to the clinic for an assessment. You know that if your child gets an unfavourable assessment there is a good chance that he or she will be sterilized or taken away and murdered. Asperger offers you a treatment that may rehabilitate the child. Obviously, with the threat of “euthanasia” hanging over your child, you will say yes to this treatment. But in modern medicine there is no way that we could consider that to be willing consent. The parent might actually not care about “rehabilitating” their child, and is perfectly happy for the child to grow up and be loved within the bounds of what their developmental disability allows them; it may be that rehabilitation is difficult and challenging for the child, and not in the child’s best emotional interests. But faced with that threat of a racial hygiene-based intervention, as a parent you have to say yes. Which means that in a great many cases I suspect that Asperger’s treatments were not ethical from any post-war perspective.

In addition, I also suspect that the research he conducted for his 1944 PhD thesis, in addition to being unethical, was highly biased, because the parents of these children were lying through their teeth to him. Again, consider yourself as the parent of such a child, under threat of sterilization or murder. You “consent” to your child’s treatment regardless of what might be in the child’s best developmental and emotional interests, and also allow the child to be enrolled in Asperger’s study[1]. Then your child will be subjected to various rehabilitation strategies, what Asperger called pedagogical therapy. You will bring your child into the clinic every week or every day for assessments and tests. Presumably the doctor or his staff will ask you questions about the child’s progress: does he or she engage with strangers? How is his or her behavior in this or that situation? In every situation where you can, you will lie and tell them whatever you think is most likely to make them think that your child is progressing. Once you know what the tests at the clinic involve, you will coach your child to make sure he or she performs well in them. You will game every test, lie at every assessment, and scam your way into a rehabilitation even if your child is gaining nothing from the program. So all the results on rehabilitation and the nature of the condition that Asperger documents in his 1944 PhD thesis must be based on extremely dubious research data. You simply cannot believe that the research data you obtained from your subjects is accurate when some of them know that their responses decide whether their child lives or dies. Note that this problem with his research exists regardless of whether Asperger was an active Nazi – it’s a consequence of the times, not the doctor – but it is partially ameliorated if Asperger actually was an active resister to Nazi ideology, since it’s conceivable in that case that the first thing he did was give the parent an assurance that he wasn’t going to ship their kid off to die no matter what his diagnosis was. But since we now know he did ship kids off to die, that possibility is off the table. Asperger’s research subjects were consenting to a research study and providing subjective data on the assumption that the study investigator was a murderer with the power to kill their child. This means Asperger’s 1944 work probably needs to be ditched from the medical canon, simply on the basis of the poor quality of the data. It also has implications, I think, for some of his conclusions and their influence on how we view Asperger’s syndrome.

What does this mean for the concept of the autism spectrum?

Asperger introduced the idea of a spectrum of autism, with some of the children he called “autistic psychopaths” being high functioning, and some being low functioning, with a spectrum of disorder. This idea seems to be an important part of modern discussion of autism as well. But from my reading of the paper [again I stress I am not an expert] it seems that this definition was at least partly informed by the child’s response to therapy. That is, if a child responded to therapy and was able to be “rehabilitated”, they were deemed high functioning, while those who did not were considered low functioning. We have seen that it is likely that some of the parents of these children were lying about their children’s functional level, so probably his research results on this topic are unreliable, but there is a deeper problem with this definition, I think. The author implies that Asperger was quite an arrogant and overbearing character, and it seems possible to me that his assumption that he is deeply flawed in assuming his therapy would always work and that if it failed the problem was with the child’s level of function. What if his treatment only worked 50% of the time, randomly? Then the 50% of children who failed are not “low-functioning”, they’re just unlucky. If we compare with a pharmaceutical treatment, it simply is not the case that when your drugs fail your doctor deems this to be because you are “low functioning”, and ships you off to the “euthanasia” clinic. They assume the drugs didn’t work and give you better, stronger, or more experimental drugs. Only when all the possible treatments have failed do they finally deem your condition to be incurable. But there is no evidence that Asperger considered the possibility that his treatment was the problem, and because the treatment was entirely subjective – the parameters decided on a case-by-case basis – there is no way to know whether the problem was the children or the treatment. So to the extent that this concept of a spectrum is determined by Asperger’s judgment of how the child responded to his entirely subjective treatment, maybe the spectrum doesn’t exist?

This is particularly a problem because the concept of “functioning” was deeply important to the Nazis and had a large connection to who got selected for murder. In the Nazi era, to quote Negan, “people were a resource”, and everyone was expected to be functioning. Asperger’s interest in this spectrum and the diagnosis of children along it wasn’t just or even driven by a desire to understand the condition of “autistic psychopathy”, it was integral to his racial hygiene conception of what to do with these children. In determining where on the spectrum they lay he was providing a social and public health diagnosis, not a personal diagnosis. His concern here was not with the child’s health or wellbeing or even an accurate assessment of the depth and nature of their disability – he and his colleagues were interested in deciding whether to kill them or not. Given the likely biases in his research, the dubious link between the definition of the spectrum and his own highly subjective treatment strategy, and the real reasons for defining this spectrum, is it a good idea to keep it as a concept in the handling of autism in the modern medical world? Should we revisit this concept, if not to throw it away at least to reconsider how we define the spectrum and why we define it? Is it in the best interests of the child and/or their family to apply this concept?

How much did Asperger’s racial hygiene influence ideas about autism’s heritability?

Again, I want to stress that I know little about autism and it is not my goal here to dissect the details of this disease. However, from what I have seen of the autism advocacy movement, there does seem to be a strong desire to find some deep biological cause of the condition. I think parents want – rightly – to believe that it is not their fault that their child is autistic, and that the condition is not caused by environmental factors that might somehow be associated with their pre- or post-natal behaviors. Although the causes of autism are not clear, there seems to be a strong desire of some in the autism community to see it as biological or inherited. I think this is part of the reason that Andrew Wakefield’s scam linking autism to MMR vaccines remains successful despite his disbarment in the UK and exile to America. Parents want to think that they did not cause this condition, and blaming a pharmaceutical company is an easy alternative to this possibility. Heritability is another alternative explanation to behavioral or environmental causes. Asperger of course thought that autism was entirely inherited, blaming it – and its severity – on the child’s “constitution”, which was his phrase for their genetic inheritance. This is natural for a Nazi, of course – Nazis believe everything is inherited. Asperger also believed that sexual abuse was due to genetic causes (some children had a genetic property that led them to “seduce” adults!) Given Asperger’s influence on the definition of autism, I think it would be a good idea to assess how much his ideas also influence the idea that autism is inherited or biologically determined, and to question the extent to which this is just received knowledge from the original researcher. On a broader level, I wonder how many conditions identified during the war era and immediately afterwards were influenced by racial hygiene ideals, and how much the Nazi medical establishment left a taint on European medical research generally.

What lessons can we learn about public health practice from this case?

It seems pretty clear that some mistakes were made in the decision to assign Asperger’s name to this condition, given what we now know about his past. It also seems clear that Asperger was able to whitewash his reputation and bury his responsibilities for many years, including potentially avoiding being held accountable as an accessory to murder. How many other medical doctors, social scientists and public health workers from this time were also able to launder their history and reinvent themselves in the post-war era as good Germans who resisted the Nazis, rather than active accomplices of a murderous and cruel regime? What is the impact of their rehabilitation on the ethics and practice of medicine or public health in the post-war era? If someone was a Nazi, who believed that murdering the sick, disabled and certain races for the good of the race was a good thing, then when they launder their history there is no reason to think they actually laundered their beliefs as well. Instead they carried these beliefs into the post war era, and presumably quietly continued acting on them in the institutions they now occupied and corrupted. How much of European public health practice still bears the taint of these people? It’s worth bearing in mind that in the post war era many European countries continued to run a variety of programs that we now consider to have been rife with human rights abuse, in particular the way institutions for the mentally ill were run, the treatment of the Roma people (which often maintained racial-hygiene elements even decades after the war), treatment of “promiscuous” women and single mothers, and management of orphanages. How much of this is due to the ideas of people like Asperger, propagating slyly through the post-war public health institutional framework and carefully hidden from view by people like Asperger, who were assiduously purging past evidence of their criminal actions and building a public reputation for purity and good ethics? I hope that medical historians like Czech will in future investigate these questions.

This is not just a historical matter, either. I have colleagues and collaborators who work in countries experiencing various degrees of authoritarianism and/or racism – countries like China, Vietnam, Singapore, the USA – who are presumably vulnerable to the same kinds of institutional pressures at work in Nazi Germany. There have been cases, for example, of studies published from China that were likely done using organs harvested from prisoners. Presumably the authors of those studies thought this practice was okay? If China goes down a racial hygiene path, will public health workers who are currently doing good, solid work on improving the public health of the population start shifting their ideals towards murderous extermination? Again, this is not an academic question: After 9/11, the USA’s despicable regime of torture was developed by two psychologists, who presumably were well aware of the ethical standards their discipline is supposed to maintain, and just ignored them. The American Psychological Association had to amend its code in 2016 to include an explicit statement about avoiding harm, but I can’t find any evidence of any disciplinary proceedings by either the APA or the psychologists’ graduating universities to take action for the psychologists’ involvement in this shocking scheme. So it is not just in dictatorships that public policy pressure can lead to doctors taking on highly unethical standards. Medical, pscyhological and public health communities need to take much stronger action to make sure that our members aren’t allowed to give into their worst impulses when political and social pressure comes to bear on them.

These ideas are still with us

As a final point, I want to note that the ideas that motivated Asperger are not all dead, and the battle against the pernicious influence of racial hygiene was not won in 1945. Here is Asperger in 1952, talking about “feeblemindedness”:

Multiple studies, above all in Germany, have shown that these families procreate in numbers clearly above the average, especially in the cities. [They] live without inhibitions, and rely without scruples on public welfare to raise or help raise their children. It is clear that this fact presents a very serious eugenic problem, a solution to which is far off—all the more, since the eugenic policies of the recent past have turned out to be unacceptable from a human standpoint

And here is Charles Murray in 1994:

We are silent partly because we are as apprehensive as most other people about what might happen when a government decides to social-engineer who has babies and who doesn’t. We can imagine no recommendation for using the government to manipulate fertility that does not have dangers. But this highlights the problem: The United States already has policies that inadvertently social-engineer who has babies, and it is encouraging the wrong women. If the United States did as much to encourage high-IQ women to have babies as it now does to encourage low-IQ women, it would rightly be described as engaging in aggressive manipulation of fertility. The technically precise description of America’s fertility policy is that it subsidizes births among poor women, who are also disproportionately at the low end of the intelligence distribution. We urge generally that these policies, represented by the extensive network of cash and services for low-income women who have babies, be ended. [Emphasis in the Vox original]

There is an effort in Trump’s America to rehabilitate Murray’s reputation, long after his policy prescriptions were enacted during the 1990s. There isn’t any real difference between Murray in 1994, Murray’s defenders in 2018, or Asperger in 1952. We now know what the basis for Asperger’s beliefs were. Sixty years later they’re still there in polite society, almost getting to broadcast themselves through the opinion pages of a major centrist magazine. Racial hygiene didn’t die with the Nazis, and we need to redouble our efforts now to get this pernicious ideology out of public health, medicine, and public policy. I expect that in the next few months this will include some uncomfortable discussions about Asperger’s legacy, and I hope a reassessment of the entire definition of autism, Asperger’s syndrome and its management. But we should all be aware that in these troubled times, the ideals that motivated Asperger did not die with him, and our fields are still vulnerable to their evil influence.

 


fn1: Note that you consent to this study regardless of your actual views on its merits, whether it will cause harm to your child, etc. because this doctor is going to decide whether your child “rehabilitates” or slides out of view and into the T4 program where they will die of “pneumonia” within 6 months, and so you are going to do everything this doctor asks. This is not consent.

The New England Journal of Medicine appears to have plunged deep into the debate on health insurance reform since Trump was elected, and in its 9th March issue has a series of articles and opinion pieces on Obamacare’s effects. This includes a piece pointing out that Obamacare expanded access to treatment for substance addiction, including opioid addiction (a big and growing problem in the US at the moment) and also a research article examining the impact of the medicaid expansion on specific health and health financing outcomes (the findings: it was broadly very positive). It also has a short research article examining the claim that the individual insurance markets have been thrown into a death spiral by the poor design of the law.

This claim has been going around for about a year now, and is generally based around the fact that some insurers have left some markets, and in some cases blamed Obamacare for their decision. For example, Zero Hedge made this claim in 2015, and the National Review took it up in July 2016. Articles discussing the alleged failings of the exchanges typically point to the withdrawal of big companies such as Aetna from some exchanges, suggesting that these companies are withdrawing because the fundamental dynamic of the exchanges prevents them from making a profit. This is important in the US context because for people earning above 138% of the federal poverty line who do not have employer-based insurance, the best and most efficient way for them to get insurance coverage is through a marketplace called an exchange, which is a special clearinghouse for selecting Obamacare-compliant insurance plans that is set up either by your state or by the federal government if your state refused to cooperate with the law. (An example of a generally well-liked exchange in a Republican-run state is Kentucky’s Kynect exchange). Obamacare’s defenders have pointed out that some consolidation is natural in markets when they change, and that new entrants or changing business practices will naturally force some businesses to fail or leave – that’s capitalism! Under this defense, the exchanges are working as intended and there’s nothing to worry about, except that in some smaller states this process may lead to a collapse of competition as only one or two insurers remain – a problem Clinton intended to fix by introducing a public provider in all markets if she won the presidential election.

The new article in the NEJM explores this issue in detail, by collecting data on all the plans that operated in exchanges from 2016 – 2017 and comparing those that left with those that remained. The authors make the particular point that once the exchanges opened the marketplace itself changed, and this had implications for insurers. They say:

In particular, the ACA’s insurance-market reforms required firms to develop and market new products that were attractive to low-income Americans who faced few access and pricing restrictions based on their underlying health status.

This means that organizations that are unfamiliar with these market conditions might struggle. They explain this as follows:

Anecdotal evidence supports the argument that the skills of particular insurers may not have been well suited to these marketplaces. Many of the exiting firms, such as UnitedHealth, have primarily covered enrollees in the self-insured–employer market, in which insurers provide administrative services and are not primarily responsible for bearing actuarial risk or for developing products targeting low-income consumers. In addition, many of the assets that have proven quite valuable in the self-insured market — such as a large national footprint that is attractive to multistate employers — may not be particularly useful in state-based individual insurance marketplaces.

They then present the results of their detailed assessment of the properties of those businesses that entered or left the market place, which they summarize in a table, reproduced as Table 1 below.

Table 1: The characteristics of leavers

This table makes clear that the insurers who left the marketplace in 2016 were offering more expensive plans with narrower networks and lower levels of behavioral health coverage; they were also much more likely to be bigger actors in the market for fully-insured people and much less likely to have experience in Medicaid markets. Overall this suggests that these companies left the exchanges not because the exchanges were flawed, but because these companies were not experienced in targeting low-income Americans who make up a large share of the individual insurance market, and having made a play at the individual market decided to get out when they were out-competed by organizations with more experience in the marketplace. The authors further note that actually a lot of the insurers active in the exchange markets are making a profit and are aggressively targeting new marketplaces – but these insurers tend to be smaller organizations with experience in Medicaid services, and don’t attract the same attention as the big employer-market insurers who failed.

This study isn’t definitive and has some limitations – for example it did not compare leavers in 2016 with historical leavers before Obamacare was implemented, and it only compared silver plans (which are the most popular but not necessarily the most profitable, I guess). Nonetheless, it gives the lie to the claim that Obamacare’s exchanges are not working, or at least suggests that they are working well enough to warrant tweaks and improvements rather than complete abolition. Once again the NEJM has shown that Obamacare’s opponents are long on rhetoric and short on facts, and that although this health care law is not perfect, it is doing okay and is certainly a significant improvement on the status quo. Let’s hope that whatever reforms proceed over the next two years will lead to improvements in the areas that are not working, and not wholesale destruction of America’s best chance at universal health coverage in half a century.

What the American people have to look forward to

What the American people have to look forward to

We’re a week away from the inauguration of the 45th President, but the Senate and House seats have changed so that the Republicans now control both houses of Congress, and one of their first actions has been to begin repealing Obamacare. They’ve been salivating over this prospect for six years and making a big fuss about it, as have all their adjutants in think tanks and conservative media, so you would think they would be ready to roll with a coherent plan. Unfortunately it appears that they don’t, and the first week of their attempts to begin the process have been rather shambolic. Since they don’t control 60 Senate votes they are trying to enact the repeal through some arcane process called reconciliation, but that is just the start of the rolling drama that is coming; Vox has an explainer about the whole process, and is running a fairly good series of articles watching as the Republicans attempt to wreck Obama’s signature achievement.

The Republicans’ first plan seemed to be “repeal and replace”, in which they would unravel all the key parts of Obamacare now but put some kind of deadline on when they would take effect, then begin working on a replacement plan in the meantime. Unfortunately this was patent madness, that they were warned about for months, which would tip many insurance markets into a death spiral and create chaos for both insurance companies and millions of insurance holders. Trump stepped on this with the announcement that repeal and replacement would happen simultaneously and soon, which is something of a problem for the Republicans since they don’t have a plan and working one up in a couple of weeks is going to be kind of challenging (Obamacare took about 15 months to happen, I think). Even more challenging for the Republicans is their lack of a filibuster-proof majority in the Senate – they can repeal the law’s components with 51 votes, but they can only put in place a replacement with 60 votes. If the Democrats decide to act in exactly the same way that the Republicans have for the past 6 years, they will prevent any replacement plan for the next two years, and unless the Republicans can hold them responsible in the mid-terms, potentially kill any future replacement. This would be a disaster for the Republicans, since they would create an insurance death-spiral with no ability to legislate a repair, and go to the mid-terms with several million people suddenly losing their insurance. Given this their choices all seem very unpleasant.

This is incredibly irresponsible politics. Health care reform has been a Democratic party priority – and part of national debate – since the 1990s, and Obamacare was passed in 2010. The Republicans have had 25 years to think about this stuff, and have tried more than 50 times to repeal Obamacare while they were in opposition, yet over that whole time they haven’t come up with a single plan that will do anything to improve health insurance coverage. One Republican even admitted that the plans they have tried to pass during Obama’s administration were only pushed because they knew they wouldn’t get passed – they aren’t serious plans. Paul Ryan has been saying the Republicans will release a plan “soon” for years, and although there are a couple of different ideas floating around out there none of them is near the level of a properly designed plan – and none were pushed during the election. The Heritage Foundation was able to scour the whole country looking for complainants in a Supreme Court case – and fight that case – to gut one part of Obamacare, but didn’t appear to have time to come up with an alternative plan that was worth putting to Congress. The Republicans have known this day is coming for at least six years and they have nothing coherent to offer the American people. We all know the reason for this, of course – Republican political ideology simply cannot produce a reform of the American healthcare system that will give more people affordable coverage, because the Republicans’ fundamental position is that government should not be interfering in healthcare markets, and it is impossible to make healthcare affordable and accessible without extensive government interference in markets.

As if that were not bad enough, their president-elect campaigned on a promise not to cut medicare or medicaid, and recently his spokesperson said that no one would lose their existing plan (a promise that has been held against Obama by Republicans for six years!) Trump has also said he likes Obamacare’s provisions on pre-existing conditions. So now the Republicans have to come up with a free market plan that somehow keeps Medicaid in place, doesn’t take away anyone’s insurance, and forces insurance companies to cover pre-existing conditions, while bringing prices down and giving individuals greater choice (the latter two points being raised by Paul Ryan recently as part of what he described as a “rescue mission” to make health care more affordable than it is under Obamacare). And if they follow Trump’s timeline they have to do it in a few weeks or months.

It’s not clear what colour everyone’s unicorn will be, but we know it will be a free market unicorn.

So what can we expect this plan to contain? It’s not clear, because there have been multiple Republican “plans” or “policies” in the past couple of years, but based on the major ones that have floated around and some of the major policy discussions we have seen, the plan will likely include some or all of the following.

  • Abolishing the mandate: The mandate is the Obamcare rule that hits people with a tax penalty if they do not take out health insurance, in an attempt to force young and healthy people to take up insurance. This mandate is key to Obamacare, since forcing young and healthy people to take up insurance will ensure that the insurance risk pools are large enough to keep costs down and keep insurance companies viable. The mandate hasn’t been as successful as its planners envisaged, probably because the plans young people are likely to choose to take up are “Bronze” plans with very poor benefits, and many young people probably don’t think they’re worth the effort of filling in forms, given the size of the tax penalty. Republicans hate the mandate and want to get rid of it but of course don’t have an alternative method for forcing people to take up health care. If you abolish the mandate but force insurance companies to cover people with pre-existing conditions then they have to raise prices for everyone else – which means the care won’t be affordable, a key goal of Ryan’s “rescue mission.”
  • Deregulating insurance markets: Trump was big on allowing insurers to operate across state lines, and most Republican plans want to see some kind of reduction of conditions on insurers. In the repeal of Obamacare this will likely involve removing the restrictions placed on plans that can be marketed on exchanges – when Obamacare was introduced, a set of minimum standards was established for insurance plans which guaranteed people buying them would get a certain minimum level of benefits, and enabled people to choose between plans that were rated as either Bronze, Silver, Gold or Platinum. By deregulating markets and the rules on how insurers market their plans, the insurance companies will be able to return to the pre-Obamacare era of selling absolutely shonky packages at a low price – which, if they’re required to offer coverage to people with pre-existing plans, is the only way they’ll cover their costs. Many Republicans also think insurance companies should be able to compete across state lines, ostensibly because this will increase competition in smaller states and rural areas where currently only one insurer operates, and also to allow more mergers. This is unlikely to encourage competition in the long-term, but will lead to large insurers merging and creating multi-state monopolies – monopoly pricing being another way to cover costs. There is no universal health coverage system in the world which operates successfully with a deregulated private market, and it’s not going to magically happen in the USA.
  • Reforming subsidies: Another aspect of some Republican plans has been to change subsidies so that they are not income-based. Currently under Obamacare anyone with income below a certain level receives a subsidy towards the cost of their health insurance, with the subsidy growing as income decreases, to ensure the plan remains affordable. This is the natural compensation for the mandate, and is one of the pillars of Obamacare. Republicans like Tom Price have proposed replacing these income-based subsidies with age-based subsidies, which means Bill Gates gets the same subsidy as a minimum-wage 61 year old labourer in Louisiana. This policy is part of a new rhetoric the Republicans are developing based on “equality of access” rather than equality of coverage. The natural consequence of this will be that poor people will decline to take up insurance, since the subsidy won’t be enough for them – especially in a deregulated market with no mandates.
  • Block-granting medicaid: As part of Obamacare the Medicaid program was expanded, with states being offered financial support to extend Medicaid to a larger pool of people (Medicaid is the USA’s free health coverage for very poor people). Republicans hate this because it’s straight-up welfarism, and the Heritage Foundation ran a successful challenge in the Supreme Court that enabled states to refuse the expansion. Unfortunately for the Republicans a lot of states – including some Republican-ruled swing states – took the expansion, and about 5-12 million people gained health coverage through it (estimates vary). If the Republicans take away this expansion they will piss off a lot of people, including people in Republican swing states that could damage them in future elections, so they need to find a way to take away the Medicaid expansion from safe Democrat and safe Republican states, and enable swing Republican states to keep it. Their answer is block-grants, in which the money for Medicaid is granted to the states but not earmarked for Medicaid only. Since some deep Republican states like Kansas and Louisiana are in big financial trouble, they can then use the Medicaid money to bail out their failing state finances, and pare back Medicaid in their states; while swing states can keep using the money for Medicaid and avoid creating a large pool of angry voters. Even then it is likely that the block grants will be smaller than the funds currently available so all states will have to cut Medicaid coverage or reduce the quality of care offered – but the Republicans don’t care because Medicaid is for poor people, so just need to make sure they don’t cut it away from so many people that it swings an election.

Any single one of these reforms in isolation would probably be enough to radically roll back recent gains in insurance coverage in the USA, but it’s likely that whatever misbegotten, evil plan the Republicans come up will have all of these reforms to some extent. This is why Republicans have started talking about equality of access rather than coverage, because if everyone theoretically has a subsidy and the right to purchase healthcare, then you can blame them if they decide they can’t afford it. In this rhetorical model they will force insurers to cover people with pre-existing conditions, abolish the mandate, deregulate the market in such a way that insurance companies can offer absolutely shonky products at inflated prices, cut subsidies so that no one takes them, and then blame poor people for “choosing” not to take up the healthcare they had “equal access” to.

It remains to be seen whether the Republicans will be able to get away with this – either because Trump takes a personal interest in a reform that actually works, and vetoes anything they offer, or because the Democrats drag out the replacement strategy until they can again win control of Congress. In any case it’s going to be fascinating to watch the Republicans try to behave like responsible adults now that they have the levers of power, even though for the past six years they have shown themselves pathologically incapable of dealing with the contradictions and challenges their ideology has thrown up.

Of course, what’s “fascinating” to those of us who live in countries with sane governments and universal health coverage, is going to be very terrifying to a very large number of poor and chronically ill people in America. Good luck to all of you!

This week’s Journal of the American Medical Association features an excellent article by Barack Obama, reviewing the implementation and outcomes of the Patient Protection and Affordable Care Act (“Obamacare”). Obviously large parts of this article were likely written by someone else, since Obama is too busy with his secret Muslim conspiracies to write a full paper, but some parts – particularly the part on why and how he implemented it – do seem to be written in Obama’s voice, which is nice to see. Vox has a brief report of the article, indicating that it is the result of a six-month Whitehouse review of the legislation and focusing on the implications of one of Obama’s recommendations (for a public option). Like most non-Americans I don’t find the recommendation of a public option to be particularly controversial or striking, so I’m not interested in revisiting it here. Rather, I’d like to briefly discuss the article’s findings on Obamacare’s achievements, take a moment to rant about what a terrible statistician Obama is, and look at some of the other conclusions he draws from his success. I will quote some parts of the article and put up one figure, but I won’t go quote too much or put up too many figures because JAMA probably wouldn’t like that. I would like to say that this is a very easy-to-read article and the choice of figures and data presentation is largely very strong – Obama certainly knows how to make a case. Also note the author affiliation: “President of the United States, Washington, DC”. Classic.

Reduction in the uninsured

In Figure 1 of the paper Obama presents the long-term trend in the proportion of Americans not covered by health insurance, and shows a huge drop after the implementation of Obamacare, from 15% to below 10% of the population. That is a huge achievement, which he states corresponds with roughly 20 million Americans receiving health insurance who would not have received it if Obamacare had not been passed. This still leaves about 30 million people without health insurance in 2015, a pretty shocking number for a developed country (in contrast, Japan has about 98% coverage and the UK about 100%). In Figure 2 Obama shows that the Medicaid expansion was responsible for a major reduction in the uninsured, by comparing the percentage drop in the uninsured in states that accepted the Medicaid expansion and those that didn’t. This drop in the uninsured increases with the proportion of people who had no insurance before the implementation of Obamacare: in a state that had 20% of its population uninsured in 2013, we see a 10% drop in the uninsured rate if the state accepted Medicaid, compared to 5% if it didn’t (these are percentage point drops, too, meaning that the proportion uninsured halved in the Medicaid state!) Obama doesn’t attempt to estimate the total number of people missing out on insurance due to the recalcitrance of the 21 states that refused to accept the Medicaid expansion, but I think the implication is obvious.

Obama’s sad statistics

Figure 2 annoys me because the straight lines shown in the plot are from an ordinary least squares regression of percentage point drop in uninsured against pre-intervention proportion of the uninsured. The straight line fit for non-Medicaid states is quite poor, because of course the relationship between percentage point drops and their starting point is non-linear. Obama would have been better served to take the logit transformation of the proportional drop, fitted a straight line model to that, and then back-transformed the resulting prediction to get two pretty s-shaped curves in his figure. I guess his article wasn’t subjected to JAMA’s usual rigorous peer review standards …

(In truth this isn’t a big deal in this case because the relationship in the data is so obvious that it doesn’t really matter how you handle it. My guess is that this figure was prepared by one of the people doing the review of Obamacare, and I would like to think that the people doing that review can do higher quality work than this!)

The three dimensions of coverage

The three dimensions of coverage

Mixed results on financial protection

Insurance is only good if it covers the services you need and offers financial protection. In health financing we talk about depth, height and breadth of coverage, which are depicted graphically in the figure above that I cribbed from an LSHTM course on financing health. Reducing the number of uninsured increases the breadth of coverage (the proportion of the population covered) but if this comes at the expense of the depth of coverage (which services are covered) or the height of coverage (the proportion of financial protection people receive) the overall benefits of the plan may be limited. Obama tackles these three dimensions in his paper, though he doesn’t use the WHO framework described in the figure above. Regarding depth, he states

Coverage offered on the individual market or to small businesses must now include a core set of health care services, including maternity care and treatment for mental health and substance use disorders, services that were sometimes not covered at all previously

Which indicates that Obamacare has forced minimum standards of coverage onto organizations that offer health insurance. This is something that people living in countries with robust universal health coverage (UHC) systems take for granted, and it’s really hard to imagine having to navigate a health insurance market where this isn’t the case – at the very least setting up a core set of covered health services reduces the risk of mistakenly choosing a health insurance package that doesn’t help you with the things you’re most likely to need it for. Obama’s language here implicitly suggests that the core package of services covered under Obamacare is an expansion of those in the previous system, but he doesn’t present any evidence that this is the case for all plans, or even in general – it could be that in adhering to these core requirements insurers have dumped some other coverage from their plans. I haven’t ever seen any research on how to assess the best services to include in a plan, or how to compare two plans that have quite different and non-overlapping benefits, so I don’t know how to assess this aspect of Obamacare (or if it can be assessed), but from the point of view of consumer protection having a guaranteed core of services seems like a good idea.

Obama's Figure 3

Obama’s Figure 3

On financial protection – the height of services – Obama makes a strong case that his legislation has been very protective. Figure 3 in the article, shown above, shows the trend in the proportion of workers enrolled in an insurance scheme that has no annual upper limit on the amount of out-of-pocket payments they must make. Out of pocket payments for health care are the main source of financial risk for individuals, and typically arise when someone has no health insurance (so must pay everything from their own money) or has health insurance with very high co-payments and deductibles, a common problem in the USA before Obamacare. Obamacare required insurers to put a cap on these out of pocket payments, and the effect on the proportion of workers exposed to unlimited financial risk is obvious in this chart. Unfortunately in a later figure we see that average out of pocket expenses haven’t changed much over time, suggesting that the annual limits that insurers placed on out of pocket payments were set high enough as to not effect the majority of such payments. To properly explore this issue we need to see data on health-related financial catastrophe, distress financing, and impoverishment due to health expenses, which to the best of my knowledge have never been adequately reported for the USA. We see some hints of this in other parts of the report, where Obama notes that the proportion of people not seeking care because they can’t afford it is down, and the average size of Medicaid debts is also down, but the picture here is incomplete. My suspicion is that a lot of healthy people have picked up bronze plans that offer them financial protection in only the most extreme cases, leaving them wearing significant costs for routine care. This isn’t in itself necessarily a problem, but to properly understand the financial protection and equity effects of the law we really need to see measures of who gets screwed by very high costs and how, rather than seeing trends in average costs.

Lessons from this policy battle

Obama concludes, unsurprisingly, that his policy has been highly effective, and I agree with this conclusion. It’s definitely not the best UHC plan out there, and even before it was rewritten by the Supreme Court and repeatedly undermined by Republicans it wasn’t a great plan, but it has achieved a lot and a lot of Americans are much better off for it. He states in the conclusion that he now wants people to accept it as the law of the land and move on to ways of improving it, but first he makes this comment about the challenges of working in American politics which gives some idea of how much of an achievement even this compromised package is:

The first lesson is that any change is difficult, but it is especially difficult in the face of hyperpartisanship. Republicans reversed course and rejected their own ideas once they appeared in the text of a bill that I supported. For example, they supported a fully funded risk-corridor program and a public plan fallback in the Medicare drug benefit in 2003 but opposed them in the ACA. They supported the individual mandate in Massachusetts in 2006 but opposed it in the ACA. They supported the employer mandate in California in 2007 but opposed it in the ACA—and then opposed the administration’s decision to delay it. Moreover, through inadequate funding, opposition to routine technical corrections, excessive oversight, and relentless litigation, Republicans undermined ACA implementation efforts. We could have covered more ground more quickly with cooperation rather than obstruction. It is not obvious that this strategy has paid political dividends for Republicans, but it has clearly come at a cost for the country, most notably for the estimated 4 million Americans left uninsured because they live in GOP-led states that have yet to expand Medicaid

Here he hasn’t gone into great detail about how the Supreme Court rewrote the Medicaid expansion part of his bill, and he has notably understated the effect of obstructionism on the Republicans, but his central point is clear: this legislation could have been better if Republicans would just have supported it, or contributed in any way at all to a constructive debate on health care. Five years have passed since the bill was first introduced to Congress, and Obama has had enough time to review its effects and write a JAMA article on it, and in all that time the Republicans have tried repeatedly to repeal it yet are still to come up with an alternative health care plan. Today they released their convention platform, and as reported by Vox it doesn’t include an alternative health care plan – in an election year. This is beyond juvenile politics, and in any other democratic polity a party that cannot come up with a coherent health policy would be treated as a joke. This is the background of Obama’s legislative efforts.

Finally, Obama makes the point that people working in health financing understand well: that UHC is about a pragmatic pathway to financial protection for everyone, not about an ideological commitment to a specific means of getting there. He says:

The third lesson is the importance of pragmatism in both legislation and implementation. Simpler approaches to addressing our health care problems exist at both ends of the political spectrum: the single-payer model vs government vouchers for all. Yet the nation typically reaches its greatest heights when we find common ground between the public and private good and adjust along the way. That was my approach with the ACA. We engaged with Congress to identify the combination of proven health reform ideas that could pass and have continued to adapt them since. This includes abandoning parts that do not work, like the voluntary long-term care program included in the law

and in this respect I also agree with him. I suspect that if the Republican party were a real political party and not a clown car, they would have recognized the importance of reform and accepted Obamacare as a practical model that protects the free market nature of the existing health system. For those Bernie dead-enders who refuse to accept compromise, nothing except a full single-payer public plan will do, and while this worked completely fine in Australia, Canada and the UK it just won’t make it in the USA, which is probably why those Berniebros find themselves in their current cul-de-sac. Obamacare is an artful example of the importance of compromise in making good health policy, and the value of practical planning over ideology. Shamefully for the Republicans and unfortunately for the country, it hasn’t been able (yet) to achieve its full promise. Obama made a few suggestions for how it can, but ultimately his particular recommendations are less important than the simple need for a return to rational policy-making by the Republicans. Whoever the next president is, she is going to want to begin tinkering with Obamacare to make it better, and hopefully the Republicans will by then have recognized that it is their responsibility to contribute positively to that process, for the good of all Americans.

I don’t see that happening, but like Obama, I can always hope …

Must hear and obey!!

Must hear and obey!!

This week I stumbled upon another one of many articles in the Guardian complaining about fat-shaming, which is apparently something society does unsuccessfully to try and force everyone to be skinny. In an interesting parallel, this week Rabbett Run has an article digging through the role of Big Tobacco in funding libertarians to talk up smoker-shaming. This “assault on smokers” is a common complaint of the Clarksons and Telegraph-letter-writers of the world: sure, smoking is bad for your health but the anti-tobacco movement has “gone too far” and now it shames and ostracizes smokers and treats them like second-class citizens (in the words of Rabbett Run’s libertarian scholar, they are parallel to the treatment of Jews in Germany!)

An interesting similarity of language exists between the anti-fat shamers and the anti- anti smokers. There is a lot of public debate about how to handle obesity, and a lot of it is denialism of varying forms:

  • Straight-up denialism, such as this book from one of the Lawyers, Guns and Money bloggers, which claims obesity is not harmful for health, the scientific research is wrong and there is nothing to fear. This crosses the political spectrum, but is usually also associated with suspicions about the diet and weight-loss industry and/or an ideology of personal health choices
  • Politically-motivated rejectionism, often feminist, which associates obesity concerns with body-normative biases and society’s obsession with controlling women’s appearance. An example of this is Fat is a Feminist Issue by Susie Orbach, which associates campaigns against overweight with historical attempts to control women’s appearance. These campaigners may deny the role of obesity in poor health (such as the Healthy at Every Size movement) or they may accept the increased risks and fall back on a logic of personal choice, but in either case these rejectionists are politically-motivated: their first concern is the ideological impact of scientific research on and public health campaigns against overweight, and this motivates their stance
  • Adaptationists, who think that it is too late to reverse trends to obesity, and/or that they are built into our society now, and so we are better off learning to adapt to these trends than try to undo them. In the medical field this manifests in a belief that we should find pharmaceutical solutions to the health challenges of obesity, rather than behavioral campaigns. This stream of thought is also common amongst anti-fat shamers and the Healthy at Every Size movement
  • First world shamers, who believe it is shameful of rich westerners to worry about eating too much when so many people in the world are starving. These deniers want us to accept that fatness is a sign of a stable and functioning society and something the world should (implicitly, usually) aspire to – similar to how some AGW denialists think that burning fossil fuels is an inevitable and necessary part of economic growth and something we should encourage the world to do, rather than trying to find alternative development paths. These first world shamers also usually ignore the fact that overweight and obesity is a heavily class-biased phenomenon in rich countries, and increasingly a problem of the poor only
  • Free choicers, who think that we should all be able to take any health risks we want if they don’t affect others, and who see obesity purely as a personal decision (i.e. fat people eat too much). Some of the adaptationists also take this view of fat as an entirely personal decision, and sadly so do a lot of public health policy-makers who want to fix the problem. Seeing obesity as a consequence of personal behavior inevitably means that public efforts to reduce prevalence of overweight will be seen as intrusive and restrictive of civil liberties, and enables these free-choicers to reframe the debate in terms of personal choices and freedom rather than the structural and social changes that are actually needed to reduce overweight. This argument is more potent when deployed in the obesity debate because it is much easier to claim that obesity doesn’t harm others
  • Skinny-shamers, the kind of hippy punchers of the anti-fat shaming movement, who see thinness as disgusting or at least present themselves in opposition to it. This fat and proud movement is distinctly political, though not necessarily associated with any party, and it is embedded in a broader cultural movement in the anglosphere towards rejecting any form of attention to appearance at all. This movement sometimes has an influence on the fashion world, especially in its attempts to redefine very thin and small models as wrong – it sometimes engages in its own form of shaming, attacking the skinny and small as wrong or ill
  • Anorexia bait-and-switchers, perhaps a subset of the fat-proud and feminist denialists, who associate campaigns against overweight with anorexia, and suggest that body-normative ideas drive young women to anorexia. In fact anorexia is a serious mental health problem not caused by social pressure or women’s magazines, and this link is spurious but it has a strong hold in popular culture, and is a powerful rhetorical device. Note that it also often relies on its own form of body-shaming, treating anorexic bodies as disgusting and accepting that they are deeply unhealthy, and often the spectrum of this body-shaming extends to women who are not unhealthy, just thin

Many of these types of obesity denialism seem to be similar in ideological composition to anti-vaccination or anti-AGW thinking. In AGW denial circles it’s common to read conspiracy theories about how the whole scare has been made up to transfer money and power to a clique; how it is cheaper and more effective to adapt than prevent; how attempts to mitigate AGW will lead to (and/or be driven by) restrictions on personal freedom. AGW denialists also often see AGW mitigation in terms of direct attacks on their personal choices rather than structural and cultural changes, for example in which they will lose individual direct choice over light bulbs and car makes, rather than seeing it in terms of industrial and community-level decisions such as changes in power generation or land use practices that no one individual can control. Arguments based on intervention in personal decision-making rather than group practices are much more amenable to conspiracy theories and assignment of nasty political motivations, and the obesity denial movement does have a fair share of such thinking.

In reality the battle against overweight and obesity cannot be won with individual changes: overweight and obesity arises only partially from personal choices, and a lot of it is driven by structural and social factors that individuals cannot change. You can’t make a decision to walk to work if your work is far away and there is no public transport; you can’t make a decision to eat healthily if there are no decent sources of fresh fruit and vegetables near your house, or if almost all the food you buy is poorly labelled and full of sugar. It’s also much much harder to stay thin if your job involves sitting for 8 hours a day, and personal decisions to do something personal to offset a structural factor are much harder to make than personal decisions that go along with that structural factor. Furthermore, cultural practices are insidious and hard to change: ideas about wasting food that come from a poorer time, types of food and eating practices are not easy to change by the time one is an adult steeped in a certain food culture. But because public health policy-makers cannot change broad structural factors outside of their discipline, like public transport and town planning, they have to focus on the things they can touch: personal behavior. This is easily construed as preaching or trying to restrict freedoms.

Of course fat shaming does happen in our societies but it’s not driven by health concerns: it’s another manifestation of a long-lasting and deeply-entrenched sexism in our society. It’s also a reflection of the fact that traditionally aesthetic values were of great importance, and people who deviated from certain aesthetic norms have been shamed for that. Compared to the way people with tattoos or men with long hair used to be treated, for example, fat shaming is nothing: no one obese ever got refused entry to Disneyland for being fat, for example. Public health campaigns do not, generally, utilize fat shaming as part of their repertoire, and the association of fat shaming with public health concerns about obesity is another example of denialism at work – and a very effective way to dampen the debate about what to do about this growing problem.

And make no mistake, it is a big problem. The continuing growth of overweight and obesity is going to have huge costs for health systems, and people who are proud to be fat now are not going to be so pleased with their personal health decisions when the musculoskeletal and cardiovascular problems start to bite in later life (when reversing the process is hardest). With the decline of smoking in the developed world, obesity is becoming the next big risk factor that will bring a wave of disease with it. Worse still, in many low- and middle-income countries overweight and obesity rates are also skyrocketing, but these countries have fragile health systems with weak financing that are not ready to manage a huge growth of chronic illness. This is a global problem, and denying it will delay the necessary steps to resolve it, leaving many countries facing an unexpected cost and health problems they aren’t prepared to deal with. Sound familiar?

I don’t think that obesity denialism is a product of Big Tobacco or Big Sugar, as was the anti-anti smoking lobby or AGW denialism. It also lacks a dimension of harm that the anti-vaccination movement carries with it (since fat kids can’t accidentally infect and kill other fat kids the way unvaccinated children can). But it has similarities with both, in terms of its scientific ignorance, the rhetorical tactics it deploys, and its blithe ignorance – or even celebration – of the problems it causes. The linked Guardian article is suggesting a need to add fat-shaming to the list of discriminatory activities that should be outlawed in Britain, which makes me think of the inclusion of “political” objections to vaccination in vaccination laws. Perhaps it’s time to start treating obesity denialism a little more seriously, before it gets a serious grip on our legislative and public health processes, making it harder for our societies to move out of a path that is ultimately not going to be good for us …