Wednesday 21 October 2015

Alcohol Concern and evidence

When I started this blog, it tended to feature swift responses to things that had been in the news that day.  These days, whether because I'm busier or lazier, I tend to be a bit slower off the mark.

A few weeks ago, Alcohol Concern published a report about trends in drinking patterns across Britain.  I found this report pretty weak and frustrating, but limited my response to a few self-indulgent tweets.  However, since then, the report has been picked up by Russell Webster, who has praised it, and so I feel that I should put down my concerns in full a bit more clearly.

First of all, I have some concerns about the sponsorship of the report.  As regular readers of this blog will know, I'm not particularly opposed to industry influence on policy, so long as that influence is clearly stated.  And influence of this kind isn't always straightforward.  I'd suggest that when tobacco companies fund the IEA, they aren't directly shaping the outputs of that organisation as much as they're making sure that a free market position has a powerful voice.  It's the same when SABMiller sponsors Demos - they know that their work is structured around ideas of 'parenting' and 'character', which suits them well, so they want to support that voice.

And so we can see the sponsorship of Alcohol Concern by Lundbeck in much the same way.  Both organisations have an interest in particular drinking patterns that are currently accepted as relatively 'normal' being labelled as problematic - Alcohol Concern because they are firm in their belief that there is no 'safe' level of alcohol consumption, and Lundbeck because they are - or were - trying to market a drug to treat a level of consumption that historically hasn't been seen as requiring any formal intervention, let alone prescribing.

So in a sense it's no surprise that one of the 'key findings' of the report was 'people who want to quit or cut down might benefit from treatments like patches or pills'.  Luundbeck, as you're probably aware, make Selincro/Nalmefene, which has within the past year been approved by NICE as a treatment for people who drink at high-risk levels but who aren't dependent.

Of course there are prescribing interventions that are useful in some cases where people want to cut down or stop drinking, but it seems odd that in a big picture style report this was seen as a major concern.  Access to medication isn't something most people in the field I speak to are too worried about.  In fact, this comes just at the point where in other comparable areas of public health - drugs and tobacco - we're seeing a possible move away from medicalised or prescribing interventions, in favour of talking therapies, peer support and innovative developments like e-cigs.

My other major concern with the findings of the report was the way it looked at alcohol consumption in isolation from other factors.

Having had a discussion along these lines with Jackie Ballard at the launch of a report on older people's drinking, I feel confident that I know why this is, and it's for the same reason that lies behind the sponsorship arrangement.  Alcohol consumption is viewed by Alcohol Concern as a problem in itself

I challenged Jackie on this point, suggesting that she was misrepresenting the nature of risk - we know that plenty of people consume at 'high risk' levels without suffering the levels of harm associated with this on average.  I didn't get anything like a satisfactory response, with Jackie simply stating that of course any amount of alcohol is harmful - even as we'd just been told that some of the people with the greatest longevity in the study had been drinking at high risk levels.

(And that raises another methodological issue with the report: using a chief executive as a specialist expert in the field to deliver findings such as 'There is possibly a view that wine isn’t dangerous; - I thought that was the sort of finding a study like this was meant to investigate, rather than leave it to armchair pundits to ponder what people might 'possibly' think.)

My particular concern in the context of the older people's drinking report was that the issue we were grappling with was how to communicate the dangers of drinking to people who can't see any ill effects of their habit.  We face the same problem with new psychoactive substances: if we go around saying that these substances kill people, it just won't sound credible to the people actually using them, who may have examples of 1,000s of instances of use without any serious harm.

And part of the problem with alcohol consumption and harm is the 'alcohol harm paradox', which suggests that although people from the highest income and wealth brackets drink the most, they also suffer the least alcohol-related harm. Although there's all sorts of issues with the data that go into these models (for example about the relevance of 'special occasion' drinking), one of the important points is that other factors exacerbate people's risk of ill health: it's not just a matter of controlling for various independent factors, as they interact and amplify each other's effects.

So when Alcohol Concern state as a key finding that 'heavy drinkers tend to be well educated' and 'people who drink care about keeping fit', they're in danger of directing our attention in the wrong direction.  It's particularly strange that they do this while noting in the very same section: “health inequalities in society also mean that alcohol has a relatively greater impact on people in lower socioeconomic groups than their higher earning counterparts”.

Of course, there's always politics about these discussions of what drinking is problematic.  The danger in being too vocal about how alcohol-related harm is most concentrated amongst lower socio-economic groups is that it reinforces that old view that working-class drinking is the most problematic, rather than domestic, middle-class wine drinking.

However, on the flip side, I have a huge fear that in a time of astonishing cuts to the budgets available for substance misuse treatment, we don't need attention being drawn to well-educated drinkers, when they don't experience the highest level of harm.

But then, that's because Alcohol Concern see a direct, unproblematic link between levels of consumption and harm.

For me, by contrast, we should start with the harm and address that, and if we need to use consumption as a proxy for harm, that's fine, but we need a lot more other data and it can only be a proxy.  So in fact this is more than a political point, it actually gets to the heart of what we're defining as being problematic about alcohol.  If it had no health effects, I think we'd be hard pressed to identify a definite problem with domestic consumption of alcohol, even if it led to intoxication (setting aside risks of setting your house on fire by cooking when drunk).

But the question of what’s problematic about alcohol consumption is a bigger discussion for a different day.

No comments:

Post a Comment