Tuesday, 2 November 2010

The nutty professor returns with bad news for booze

 We are obsessed with league tables.  However arbitrary, however unaccountable, however biased - we love them and cannot get enough of them.  There are league tables for schools, universities, hospitals, and even, lately, a league table of how vulnerable local authorities might be to Government spending cuts.

And now a new league table of harmful drugs has been published in The Lancet by the Independent Scientific Committee on Drugs, entitled 'Drug harms in the UK: a multicriteria decision analysis'.  The study claims to classify drugs on the basis of the harm that they do  to users and others and, according to the scientists' criteria, alcohol has the dubious honour of finishing top of the pile.  It is some way ahead of Class A drugs heroine and crack cocaine, and a proverbial mile ahead of the likes of metamfetamine, cocaine and tobacco (see table below).
Drugs ordered by overall harm scores, showing the separate contributions to the overall scores of harms to users and harm to others
The weights after normalisation (0–100) are shown in the key (cumulative in the sense of the sum of all the normalised weights for all the criteria to users, 46; and for all the criteria to others, 54). CW=cumulative weight. GHB=γ hydroxybutyric acid. LSD=lysergic acid diethylamide.
Notably, this research has been conducted by a team that includes Professor David Nutt, the former government scientific adviser.  David Nutt was dismissed last year by the then Home Secretary (now Shadow Chancellor), Alan Johnson, for criticising the Government for ignoring advice given by his Advisory Council on the Misuse of Drugs on political grounds instead of scientific evidence.  Professor Nutt claimed that ecstasy and LSD were less dangerous than alcohol (which appeared in fifth position back then) and should be downgraded.  The Government, not wanting to be seen to be soft on illegal drugs, demurred.

Now, however, David Nutt is chancing his arm once more.  Almost exactly one year ago, he  put his suggestions to a Labour Government that after nearly thirteen years in office was out of steam, out of ideas and bereft of political capital.  This was, moreover, a Government that had governed by opinion poll and, particularly under Gordon Brown, licked the boots of the Daily Mail on law & order.  Did Nutt really think this was a propitious time to suggest a radical reappraisal of this country's drugs policies?
Professor David Nutt

Of course it was not, and so it proved.

And here Professor Nutt finds himself, faced with a new set of ministers - a novel coalition ministry, no less - imbued with reforming zeal and embarking on a programme more radical than most could have imagined.  This is particularly true in the case of health, where Andrew Lansley plans to make the most significant changes to the NHS since its foundation.  Moreover, the Conservative-Liberal Democrat agreement pledged to "reverse the substantial erosion of civil liberties under the Labour Government and roll back state intrusion."

This does not imply that the coalition partners wish to downgrade drug classifications, but it does imply that there is now an atmosphere more conducive to the sort of liberal thinking that Nutt advocates.  On top of that, the Programme for Government says the coalition "will not permanently ban a substance without receiving full advice from the Advisory Council on the Misuse of Drugs."  Again, this does not imply that declassification is on the agenda, but it offers reassurance to the precise body rebuffed by Labour last year that it will be properly consulted.

Overall, this is a far more propitious atmosphere for David Nutt to try again - only this time, he and his colleagues have been more radical still, and it makes me uneasy.  Whatever the scientific evidence - and I am in no position to doubt it - there are some not insignificant misgivings to be had about this announcement.

First, league tables are interesting as a rough and ready starting point for debate but not as conclusive evidence for action.  The table is essentially a product of its criteria and, as the widely vacillating international university league tables testify, slightly altered criteria can produce a strikingly dissimilar outcome.  Nutt et al recognise this flaw themselves as they write:
To provide better guidance to policy makers in health, policing, and social care, the harms that drugs cause need to be properly assessed. This task is not easy because of the wide range of ways in which drugs can cause harm. An attempt to do this assessment engaged experts to score each drug according to nine criteria of harm, ranging from the intrinsic harms of the drugs to social and health-care costs. This analysis provoked major interest and public debate, although it raised concerns about the choice of the nine criteria and the absence of any differential weighting of them.
They go on to claim that these concerns have been addressed by employing a multicriteria decision analysis (MCDA) approach, something we are told has been "used successfully [in]...complex issues characterised by many, conflicting objectives - e.g. appraisal of policies for disposal of nuclear waste."  To a layman like me, that doesn't sound like a perfect analogy.  Even if that is irrelevant, the fact remains that their league table criteria remains very much open to interpretation.  After all, alcohol has managed to jump from 5th to 1st this year on the basis of different criteria, just as my alma mater the University of Bristol had climbed from 34th last year to 27th this year in the QS World University Rankings, whilst only coming 68th in the Times Higher Education World University Rankings 2010-11.

Furthermore, MPs and public sector professionals regularly criticised the  Labour Government for its use of league tables in schools and healthcare.  The House of Commons Public Administration Committee questioned in 2002 whether NHS league tables were of any practical use, and headteachers disputed the value of school league tables as a marker for parental choice.  In November last year, ministers and health regulators questioned what was a more reliable form of hospital ranking, the Dr Foster Intelligence report, or the work of the Care Quality Commission. The big problem?  With any league table, there'll always be disagreement on the objectives and the formulating criteria.

Secondly, Nutt et al have said "alcohol" is the most harmful drug, without discriminating between its many forms, e.g. wine, beer, cider or spirits.  They do, however, differentiate between similar drugs such as metamfetamine and amfetamine.  This is not a trivial observation.  There are significant differences between types of alcoholic drinks on a socio-cultural level.  To be fair, David Nutt has been quoted as calling current alcohol taxation "completely inappropriate", with duty on strong cider (associated with youth drinkers and alcoholics) a fifth of that on wine (seen as more refined, communal and carrying antioxidant health benefits).  It is a major omission not to bring this out in their report.  I can accept that drinks like super strength cider and beer, drunk in their most depraved circumstances, are just as harmful, if not more, than many drugs.  But a sensible bottle of claret at 12.5% shared between friends at dinner?  I think not.  It is unfair and misleading to lump the two together as has been done in the report.

My third concern is that ranking illegal drugs in this way could potentially send out entirely the wrong message: that illegal drugs are dangerous only relative to each other, not absolutely.  The language and presentation are deleterious because Nutt et al rationalise one drug against another, internalising "good" and "bad" within a single shopping list.

Nutt et al are not being intentionally irresponsible.  Their paper is littered with caveats surrounding the criteria and weighting methodology.  It is also "note[d] that a low score in our assessment does not mean the drug is not harmful, since all drugs can be harmful under specific circumstances."  Yet this is not brought out in any of the media coverage and nor will the vast majority of people who glean information from the report's coverage read the actual report and notice that sentence right at the end.

So I think it is valid to fear the unintended consequences.  This league table could plausibly  be perceived as, at best, a rational, scientific 'shopping list', akin to guides of what foods have the most Vitamin C, or the lowest GI value, or, at worst, a kind of game to be exploited.  In December 2009, the NHS released an alcoholic drink units tracker application for the iPhone in order to help stop people overindulging.  It is a clever use of behavioural nudge theory, a good tool and one I continue to use to this day.  However, the app was also found to encourage exactly the binge drinking it was designed to avert.  An iTunes review described it as an "awesome game" and gave it 5 stars.  “This is the best game ever,” it continued. “Gonna try and outdo last night’s score on Friday night."  There were many more, one even suggesting an online league table of high scores.

Who is to say that a drugs league table might not, in time, develop a similar sort of following?  The Government would certainly be wise to thank Professor Nutt and his colleagues for their work but to keep the research at arm's length, so as not to legitimise the concept, like the NHS did with its drinking app.

All things considered, the report's biggest failing is sloppy communication of objectives and a misguided media message.  Considering the axe Professor Nutt has to grind since the events of last year, this is not wholly surprising.

Eighty-seven years ago, the writer Rudyard Kipling said, "Words are, of course, the most powerful drug used by mankind."

Kipling was no drugs expert or scientist, but he was not a bad authority on words.  And it is on the words - or the presentation - that this report will have its greatest effect, not the science it contains.


  1. I cannot help but wonder (and basely speculate!) if the scoring has been skewed by the panelist's personal experiences. Each panelist is far more likely to have experienced some kinds of negative impact relating to alcohol whether through friends, family or even themselves.

    However, each is very unlikely to have experienced a friend or family member going through the trials of abuse of many of the other substances scored, instead observing and studying such cases from a distance (therefore much less close to home).

    Just a thought.