At the end of October, just ahead of releasing the 2016 statistics, the Office for National Statistics (ONS) updated the definition used for recording alcohol specific deaths.
The change in definition meant the exclusion of deaths that were only “partially attributable to alcohol”.
The ONS anticipate that the impact of this change will be a reduction in recorded alcohol deaths (a reduction of 18.7% in men and 24.6% in women). The reason for the change is to ‘harmonise measures’ across agencies, including Public Health England.
So what? Surely robust and harmonised recording can only be a positive for the published statistics?
The changes cause us concern because statistics are not produced or used in a political vacuum, these new figures are inconsistent with other Public Health messages. We could also argue, do not represent the reality of the damage caused by alcohol in the communities where we work.
At a time when budget cuts to drug and alcohol services are being seen, and there is an increasing voice of concern about the impact of these, it is imperative that the changes in definition are understood to be the reason for the decline in the published statistics.
Until this redefinition, alcohol specific deaths in the UK were consistently high and across comparable populations (e.g. people who use drugs, people in contact with the criminal justice system) mortality rates are increasing.
We need to be vigilant that this ‘drop’ is not used to mislead the public on the pressures that drug and alcohol services are facing and the impact that austerity is having upon people who need these services.
This is compounded by a lack of national strategy around alcohol, the recent drugs strategy was weak in its discussion of the harm caused by alcohol, and links with drug treatment.
The redefinition however harmonising with the statistics is actually in stark contrast to the public health messages that alcohol is related to 60 different health conditions including seven types of cancer, and raising the profile of the social and familial difficulties it causes.
The exclusion of deaths partially attributable to alcohol therefore makes no sense in the context of the known harms that it causes. To not consider partially attributable deaths is to not recognise the impact of alcohol upon these wider health needs. Finally, the impression of a reduction in the number of deaths contradicts other information available about the harm caused by alcohol to individuals and families across the country.
The Independent reported this week that more than 1 million hospital admissions each year are attributable to alcohol, whilst Pryce et al (2017) estimated that 595,131 adults require specialist intervention for alcohol use and that 222,077 children are living with an alcohol dependent adult.
Public Health England’s 2017 report stated that 4 out of 5 alcohol dependent adults do not access treatment services.
This means that there are 476,105 adults, with approximately 89,000 children that will not access services. This level of unmet need is clearly not congruent with a reduction in deaths.
If and how the new figures that have now been published will be used is yet to be seen, and the actual impact of the redefinition may not amount to much.
However given the ease with which statistics can be misconstrued, used and hidden behind, this is something to watch.