A strong headline.
If it isn’t true, I can easily be sued for libel. I’m not expecting to be served legal papers any time soon, and that’s because of two news stories published today.
The first is from the hateful, fear-mongering Daily Mail. Under the headline ‘‘Meteoric rise’ in alcohol-related deaths’, the Mail gives a summary of Sir Ian Gilmore’s speech at a conference yesterday hosted by Alcohol Concern. In this speech he cites a ‘meteoric’ rise in deaths by liver disease, and we are told that alcohol-related hospital admissions are at an all-time high. The article also mentions a 2011 study showing that 30% of boys and 25% of girls claim to have been drunk in the last thirty days.
This all seems very clear. Except it isn’t.
Also today, Public Health England announced that it will be changing the method of alcohol-related hospital statistics following acknowledgement that the figures quoted yesterday are misleading. Hospital admissions are broken down into primary and secondary causes. If you get so drunk you have alcohol poisoning, alcohol is your primary cause of admission. If you’re admitted with liver disease or high blood pressure – which could be caused partly by drinking, as well as other factors, alcohol is a secondary cause of your admission.
Even if you don’t drink.
It goes beyond that – I’ve written here before about how if you have an accident or injury, and you have had a drink, your admission is alcohol-related even if that drink did not – could not – have been relevant. If you’re having a glass of wine in a restaurant and the roof caves in on you, for example, your injuries are alcohol-related.
So the body that releases the statistics is recalculating them because they are misleading, splitting out primary and secondary causes more clearly. Alcohol Concern and Ian Gilmore know this, even as they continue to cite these statistics.
But today’s report reveals something even more extraordinary. Because even if you think the stats are accurate and true, as I’m sure Gilmore and Alcohol Concern do, according to the people who compile them, you cannot use them to suggest that alcohol related hospital admissions are increasing – as Gilmore and friends frequently do. Here’s what a spokesperson for Public health England has to say:
Much of this increase is believed due to improvements in diagnosis and recording… these improvements mean that while recent estimates are likely to be a better reflection of the comorbidity [secondary disorders] associated with alcohol, estimates from earlier time periods are not directly comparable as they will have underestimated the number of secondary conditions related to alcohol. [My emphasis]
So, depending on whether you are pro- or anti-drink, either:
Gilmore and Alcohol Concern are talking bollocks because the official figures overestimate alcohol related hospital admissions
or
Gilmore and Alcohol Concern are talking bollocks because the official figures show an increase only because of improvements in measurement, not because of changes in behaviour.
Either way, these people know about this. They know they should not be using these figures to claim a rise in alcohol related hospital admissions. But they do it anyway, wilfully misleading the nation.
In addition, Gilmore and Alcohol Concern repeatedly avoid the medical fact that only around 37% of liver disease is primarily caused by alcohol – it’s also caused by Hepatitis C and obesity. They never refer to Britain’s rising obesity epidemic as a possible cause of rising liver disease. It must be alcohol consumption, even though that is declining long term.
Oh, and those figures above talking about the percentage of kids drinking? What the Mail refuses to tell you is that the survey from which they were taken showed a REDUCTION in underage drinking. That’s why they don’t tell you what the figure was a few years before.
We are being lied to. Tell everyone.
Cheers Pete, for bringing this prickly issue to our attention. It doesn't surprise me one jot about the points Alcohol Concern have raised. I have done articles both in the local press and on my blog about this self-righteous , puritanical , holier-than-thou and blinkered group.
Well done on highlighting this devious bunch, who continue to dupe and mislead us – well,attempt too!
The first things they want to answer for bad health is smoking or drinking then linking everything up with that afterwords. Instead of looking at the real reasons it is an easy short-cut which the Government uses because it is easy. I have had this first hand. Defy this by having a beer.
Does anyone know what would happen to the highly salaried jobs of the good people of Alcohol Concern should public outrage about alcohol ever die down?
This couldn't possibly be seen as a motivation for them to lie and obfuscate, could it?
Pete, have a listen to the episode "Underage Drinking" of the R4 programme "More or Less" from 4th Oct. It's available as a podcast. Amongst other things, it points out that a child admitted because he has "drunk bleach" will appear in the stats for children admitted for being drunk because the software that scans the records simply looks for the word "drunk."
http://downloads.bbc.co.uk/podcasts/radio4/moreorless/rss.xml
Good work Pete.
The consultation is actually arguing that the statistics are not comparable because when you were just pulling random sets of notes for audit and deciphering the scrawls, you probably didn't get a complete picture. It acknowledges that asking the same question to databases generated by hospital IT tends to get higher returns.
The actual method of producing alcohol statistics isn't really changing. Basically, every disease is scored on the probability of alcohol being a causative factor. Some things, like alcohol intoxication and alcoholic liver disease have a probability of 1 i.e. alcohol is always causative. In Pete Brown's example, if you had the less specific diagnosis of all-cause liver disease, the probability is 0.37. It could also be applied to less obvious conditions like fatal car crashes or someone self-harming – in a proportion of cases, alcohol is a root factor.
The system that is currently used is capable of listing 20 diagnoses, so if you include all the low probability diseases in your all-comers alcohol disease figures, it gives an artificially high number – which the alcohol industry reasonably considers unfair, because it leads to graphs like this one, with a steep rise in admission related to alcohol.
Yes, you may be in with a heart attack, due to your high blood pressure, which may be related to your alcohol intake. However, the clustering of poverty, smoking and obesity make laying this as an alcohol related disease a bit tricky and would give a low probability. Times a low probability by a high number (such as heart attacks) and you will be bumping up the figures in enough areas to inflate reality. If you restricted your numbers to only those secondary causes that were more obviously alcohol related (e.g. in this case, specific diagnoses such as alcohol induced arrhythmias or cardiomyopathy), you would find the numbers drop down.
So what is planned is to analyse based on an agreed set of alcohol related diseases (diseases with higher probabilities, essentially). What happens if you do that? Well, you get this graph, which shows a persistent rise in alcohol related admissions. The 2010/11 figures are quite close to that first graph, so what is directly attributable to alcohol at present is unchanged. It is the start of the decade that is drastically increased, suggesting the problem was always much worse than the statistics suggested.
There simply has been a rise in alcohol related hospital admissions.
So, is this rise less than meteoric? Well, a rough 40% increase in ten years on top of those increased figures from the start of the decade seems quite significant to me. That's before we get into a debate about whether admissions is really the right figure to track, or the implications on length of stay and actual death and disease impacted on by alcohol.
Oh and that medical fact about only 37% of liver disease being caused by alcohol? You'll find it plastered all over alcohol related health material, because the next highest contributor is liver cancer at only 22% (and is itself a condition associated with alcohol intake). Fatty liver disease (not synonymous with obsesity), only 6%. Viral hepatitis? Sure, a big player on the world stage – but a paltry 2% (rounding up), in England. Also, all these numbers are actually deaths, not disease and come from this report.
This is a version of a reply to this blog post, via reddit (edited for character limit).
I have worked within the alcohol field for quite some time now, and the evidence is there regarding the rise in alcohol related harm and health issues. I think that those that chose to ignore this are either not directly affected enough, in denial, or working for the alcohol industry. Sir Ian Gilmore is not scare mongering but highlighting what those working in treatment services, hospitals and others public services are seeing every day.
Generally Sceptical raises some interesting points and, when you get into the realm of how to interpret statistics, it can be easy for the layman to get 'lost' in the data!
My question(to Generally Sceptical or anyone else who knows) is, at what level does a condition become 'alcohol related'? Obviously a probability of 1 would be included and I could see that probabilities of 0.5, 0.33 and 0.25 could be included, but what about probabilities 0.2, 0.15, 0.1 and lower?
To properly gauge the statistics we really do need a definition of what 'alcohol related' actually means.
One final point; the graphs linked to by Generally Sceptical also show that the rate of increase is slowing down. This MAY be reflective of the fact that overall alcohol consumption has been going down since 2004 and is indicative of a lag in cause/effect being about 10 years which seems reasonable as these are mostly chronic conditions.
For Peter Allen, you can find the methodology and subset of conditions here http://www.lape.org.uk/downloads/AlcoholAttributableFractions.pdf
I won't pretend to have read all the methodology, but you can skip to table 17 on page 44 to see a table of conditions and alcohol attributable fractions.
The plan, as far as I can see, is to produce the statistics with a set of figures for primary diagnosis/cause of admission only, as well as the set with secondary causes with high fractions. I can't find any documentation to say what exactly will count as a high fraction. I would assume, like the conditions themselves, they are likely to be set by committee consensus.
My thanks to Generally Sceptical for the link! (My brain hurts now!)
It is obviously a very complicated area and it is good to see how it is being academically pursued. But, the longer term problem is, for those of us without statistical training, finding out the proper interpretation of the figures. Essentially, the anti alcohol lobby will paint as bad a picture as possible and the pros will attempt to do the same from their perspective with the real truth being lost in the fog of 'war'!
Back to the original point of the blog post, though. For those of us with a pro alcohol leaning it is good to see someone stand up to the anti brigade…so, well done Pete!
There are still some shocking anomolies when it comes to hospital admissions – a few years ago, my partner was admitted to the local hospital with severe stomach pains, causing vomiting etc etc. As these had been building over a few days, it was at least a week since she had last had any alcohol. However, her admission sheet listed primary cause of admission as Alcohol as she was wearing a beer festival T-shirt! When we pointed out this was clearly nonsense, we were told this was NHS policy!
There is a problem with alcohol related illness, but the reporting is so murky and so skewed I wouldn't have thought anyone really knows what's happening.
Lives are ruined and people die from alcohol abuse, however you look at the statistics. How would you make money if the alcohol industry went out of business Pete???
Oh look, another anonymous poster.
Anon, I have never, ever denied that people's lives are ruined by alcohol. I know this from first-hand experience. What I am saying is that the anti-alcohol lobby is lying about the scale and the extent of the problem – and NHS and government figures say so.
What makes me very angry about this is that by proposing measures that punish everyone and seek to demonise alcohol and restrict overall availability, they (a) reduce our collective understanding of alcohol rather than increasing it and (b) fail to help those who really need it.
Alcohol Concern and their ilk seek to restrict access to alcohol for everyone and assert that there is a direct link between the amount we drink and the problems that result. This is false – figures show that the more affluent you are, the more likely you are to drink alcohol, and the less affluent you are, the more likely you are to suffer from alcohol related harm. It's not the fact that people drink, but how and why they drink. And as long as anti-alcohol bodies continue their lies and scare tactics, no one is examining the true causes of alcohol abuse, and this is increasing the suffering of those in trouble.
In answer to your question, I'd simply write about something else. And if you DARE to imply that my opposition to Alcohol Concern and their lies is driven by simple selfish interest in protecting my income as a drinks journalist, at least have the fucking balls to put your name to that accusation.
Hi Pete
Interesting post. However I'm not sure I 100% agree. I a) work for a drug/alcohol charity and b) am a massive beer fan! So I do honestly feel I have a foot in both camps. I was at the Alcohol Concern conference and I also go to beer festivals, use ratebeer.com etc and check out blogs like yours.
I must admit I don’t feel totally qualified to comment on the statistics, especially in light of what the commenter above says. It does sound like it was rash to use the stats as a point of comparison if PHE have expressly said they cannot be used in this fashion. However
• We DO have a problem with our alcohol consumption in this country. From 1992 to 2008 alcohol deaths have more than doubled from 4023 to 9031. This trend cannot be healthy. Since 2008 it’s been up and down.
• Almost all other Western European countries alcohol deaths are going down, including France and Italy who traditionally had very high levels, and the UK (particular Scotland) are approaching the Eastern European bits of the graph which is not good news as Russia, Ukraine have sky high rates of problem alcohol consumption! We are not fitting into the general ‘direction of travel’ for W Europe.
• ‘It's not the fact that people drink, but how and why they drink.’ Wrong – it’s all three. You correctly identify that richer people are more likely to drink more and also less likely to develop alcohol problems. Totally true – social ‘capital’ (employment, housing, family, friends, income) all play a key protective role in preventing people developing serious drug or alcohol problems. But the amount of ethanol people consume remains ‘the problem’ and the first indicator that something is going wrong. It doesn’t matter if it’s the finest Petrus or White cider, the human body does not differentiate even if the palate does. You can still be completely ruining your health (how many people know for instance that alcohol causes cancer??) and making life a living hell for your family and be financially solvent, holding down a job and maintaining a semblance of normality to the outside society.
• The idea that health organisations campaigning on alcohol are all anti-alcohol and run by teetotal party poopers is a complete myth! I have been to the pub with lots of these people and seen them enjoy a drink. Some people may preach but education and campaigning on alcohol is totally sound public health stuff that will save lives AND save our state lots of money.
• If you look at the national debate the players are roughly 1) the public 2) the drinks industry 3) public health people. The drinks industry obviously have a huge stake in telling us we don’t have an alcohol problem as a nation as they are rolling in it to a quite obscene extent and frankly want people to be addicted to alcohol as they make the best customers. Their message is that problem drinkers are just the tiny minority of people sitting in parks with dogs drinking white cider and that everyone else is fine. This is not the case. So they will never say ‘drink less’ unless in some flimsy lip service way. The public (including me, I love beer and would happily have a few beers every day of the week, though I try not to) have an interest in self-denial and not thinking of alcohol as a drug. So they don’t want to hear the message to drink less. Lastly the public health people are telling us all to drink less and to drink more safely. What is their motivation? Well you could claim that this is all part of a ruse to keep themselves employed. But if you look at the evidence this is not the case. Ian Gilmore is (or was) a liver specialist – he has been on the frontline of treatment with very ill people. Whatever way you look at it alcohol is a massive public health issue which harms our health AND costs us a lot of money in treatment, A&E and policing. It just makes sense for any country to tackle big public health issues. It would be irresponsible not to I believe!
Sorry for the essay, and like I say I LOVE beer, and am in no way anti-alcohol or a killjoy. But our society has completely lost any idea (if it ever had it) that alcohol is a drug and is something which is causing large harms to our society. Because of its social acceptance, advertising and, let’s be honest, the fact that alcohol is fun and tastes amazing we are not being honest with ourselves. I agree – let’s not demonise alcohol, but let’s acknowledge it is a drug and we’ve got to do better at dealing with it. I’m more than happy to send over some links if that is of interest.
Cheers, and thanks for all the blogging
My husband died of cirrhosis caused by alcohol. Our children were 12 and nine. It's taken me some years to accept drinking as a social norm again – however, I want my teenage sons to live a normal life and that involves fitting in a culture that enjoys drinking alcohol. There is nothing wrong with drinking as such, but I welcome research into understanding the harm it can do and the work Ian Gilmore is associated with – it's not just about government stats and scare mongering.