NHS Alcohol Statistics: the war of numbers

The launch of the Coalition Government’s new alcohol harm reduction strategy coincides with another change to the way in which the statistics of alcoholrelated admissions to NHS hospitals are collected and presented. On this occasion the result is that the number of admissions appears to have declined rather dramatically.

The number of admissions is, of course, an issue of fundamental importance because these figures are one of the key indicators used to measure the size and significance of the national alcohol problem, and they are one of the determinants of the estimates made of the size of the economic burden placed by alcohol on NHS budgets and on the public purse. Clearly, the higher the number of alcohol-related admissions to hospitals there are, the greater the economic and social burden of alcohol appears to be. And, of course, the greater the problem is perceived to be, then the greater the pressure for policy measures to ameliorate it.

The Coalition’s New Improved Figures

The change in the Coalition Government’s approach to the numbers issue became apparent when, in advance of the launch of the new alcohol strategy, Prime Minister David Cameron made a highly publicised speech in which he promised to tackle the scandal of the binge drinking culture and the burden it places on society, and on the health service in particular.

Mr Cameron said:

“In 2010/11 alone there were 200,000 hospital admissions with a primary alcohol-related diagnosis, 40 per cent higher than in 2002/03. The number of patients admitted with acute intoxication has more than doubled to 18,500 since 2002/03.” Mr Cameron went on to say that the figures show an ever-growing bill to the NHS, which currently stands at £2.7bn a year, including £1bn on accident and emergency services. £2.7bn equates to £90 for every taxpayer. This is part of a wider cost to society from alcohol of between £17 billion and £22 billion per annum.”

Mr Cameron’s reference to the figures for 2010/2011 may have been a mistake, because the figures he quoted are more probably for 2009/2010. But the significant point is that the figure of 200,000 is a fraction of the figures that had previously received much media attention, the headline claim being that alcohol-related hospital admissions in England had passed the million mark.

What the figures mean

The difference between 200,000 and over a million is the difference between admissions where there was a primary diagnosis of a disease or a condition wholly attributable to alcohol, and the much larger number of admissions in which alcohol was recorded as a secondary or accompanying factor. (see graph)

It appears that the Coalition Government is now minded to revise the way the alcohol admissions statistics are presented to show only those where alcohol entered into the primary diagnosis. Under the new Public Health Outcomes Framework, the North West Public Health Observatory will undertake a consultation on the methods used to measure alcohol-related admissions, and the definition adopted will be informed by the results of that consultation. However, the fact that Government Ministers are already referring to the lower figures suggests that the decision to change the system has already been taken.

What was wrong with the old figures?

Headlines about a million plus alcohol admissions to hospitals are indisputably very bad publicity for the alcohol industry, and it is understood that industry groups have been making representations to Government to change the way the figures are presented.

However, the old method of presentation had also been criticised by some more impartial commentators who argued that changes in the way alcohol admissions were recorded made comparisons with earlier years worthless and claims about a millionplus admissions extremely questionable.

The key argument is that the method of coding hospital admissions has encouraged NHS hospitals artificially to inflate the number of admissions related to alcohol.

The issue is that the million-plus figure is based on taking into account all the conditions recorded for each patient, not just the one for which the patient is actually admitted. Nigel Hawkes, the science journalist and founder of the Straight Statistics website, points out that many seriously-ill patients are admitted with several conditions: a lung cancer patient might, for example, additionally suffer from high blood pressure. The high blood pressure is not the condition for which the patient is being treated and it is not responsible for the admission to hospital, but it will also be recorded on his or her notes, and ‘coded’ by clerks. Lung cancer is not an alcohol-related disease, but high blood pressure is, with alcohol being implicated in around one third of cases. Hence, of every ten admissions to hospital of this kind, around three will be recorded as alcohol-related, even though alcohol was not the cause of the admission.

Hawkes writes:

“This might not matter much, beyond greatly exaggerating the numbers harmed by alcohol, if it were not for the variation in coding practice over time and between hospitals. The introduction of payment by results by the NHS in England and the publication by health analysts …. of death rates for individual hospitals have encouraged hospitals to increase the number of codes attached to each patient.”

This is because, if a hospital can show that a patient who died in its care suffered a whole range of conditions on admission, then that death will count for less in the calculation of its Hospital Standardised Mortality Ratio (HSMR) and its official performance rating will improve commensurately. Health analysts have worked out that the average number of codes per patient in the NHS in England rose from just under 3 in April 2005 to almost 4.5 by December 2010, an increase of 50 per cent. If the number of codes increases, then so will the number of ‘alcoholattributable’ codes. So, Hawkes argues, a rise of 50 per cent in alcoholattributable admissions over that period can be accounted for simply by ‘coding creep’.

However, while there is clearly substance to Hawkes’ claim that changes to the coding system have exaggerated the increase in alcohol admissions over time, there is an even clearer danger that counting only those in which alcohol forms part of the primary diagnosis will result in a large underestimate. There is a range of conditions which are not defined in terms of alcohol but which are partly attributable to it, and which, therefore, contribute to a patient’s admission to hospital. To exclude these would be to engage in the opposite kind of misrepresentation.