The Health and Social Care Information Centre (HSCIC) published new data on Wednesday 1 October 2014, which shows a steady 37% increase in alcohol-related hospital admissions in England over the last 10 years (236,770 in 2003/04 to 325,870 in 2012/13).

This steady increase in admissions is a cause for concern, particularly for those working to reduce alcohol harm. 

Why are alcohol related hospital admissions rising?

It is still unclear exactly why admissions are continuing to rise considering falls in recorded alcohol consumption overall. One explanation may be that serious harm from alcohol often takes some time to be visible in health statistics (1). Also the so-called ‘harm paradox’ (2) may play a part in explaining the unevenly distributed social impact where less well-off people are more hard hit.

The data provides a comparable 10 year time series, which is a useful way to look at how admissions have changed over the last decade. But the way that the data has been recorded has changed and we must consider this before drawing any conclusions. 

How are alcohol-related hospital admissions calculated?

Alcohol-related admissions are calculated using the sum of the contribution alcohol harm is making across all hospital episodes. It is not a count of individual admissions. 

Importantly this new data has been published based on a revised methodology for recording alcohol-related admissions. The previous ‘broad’ measure has been supplemented by a new ‘narrow measure’.

How is the new ‘narrow measure’ different?

The ‘broad’ measure is more likely to overestimate the role of alcohol in causing admissions, while the ‘narrow’ indicator is more likely to underestimate the role of alcohol.  It is thought that using them together presents a more accurate picture than just using one of them.

Pros and cons of the new ‘narrow measure’

Pros:

  • The new, narrow indicator is higher in 'specificity' and is better for measuring change over time or differences between areas as it is less sensitive to variation in coding practices.

Cons:

  • The new narrow indicator may lack 'sensitivity' and underestimate the impact of alcohol on hospital admissions.

Key alcohol statistics using the revised method

Here are some key findings from HSCIC (3) using the revised measure of alcohol-related hospital admissions:

  • In 2012/13, there were an estimated 325,870 admissions based on the narrow measure.

Of these:

  • 32% (103,160) were for wholly attributable conditions (where one admission equals one person admitted), including over 16,000 admissions for alcoholic liver disease.
  • 68% (222,700) were due to conditions which were categorised as partly attributable conditions – both acute and chronic (here, each count of an admission is made up of several fractions of individual admissions, based on the role played by alcohol harm). This includes over 72,000 admissions for 6 different types of cancer.
  • 62% of admissions involved men (202,580) compared to 38% of admissions involving women (123,280). This reflects a higher level of harmful drinking among men compared to women overall.
  • The rate of alcohol-related admissions varied regionally from 820 per 100,000 population in the North East to 490 admissions per 100,000 population in theSouth East. The average for England is 610 alcohol-related admissions per 100,000 population.

Public Health England opinion

According to Clare Perkins and Matt Hennessey at Public Health England:

“The new supplementary indicator provides a narrower measure of alcohol harm that is less sensitive to the changes that have occurred in coding over the years and therefore enables fairer comparison between levels of harm in different areas and over time.

"It is also more responsive to change resulting from local action on alcohol. However, the original indicator is a better measure of the total burden that alcohol has on community and health services.”

An important note on Alcohol Attributable Fractions

It is important to note that these figures are calculated as alcohol-attributable fractions of hospital episodes, not the number of individual people or admissions due to alcohol. ,

Fractions are used “as a measure of pressures from alcohol on health systems”, and take into account things like age and gender.

For example:

  • If the admission is as a result of alcoholic liver disease, this can be wholly attributable to alcohol, so will be counted as 1 admission.
  • If the admission was as a result of an assault, which has been estimated to be attributable at least in part to alcohol, it will count towards part of an admission. In this case it will be 0.27 of an admission as epidemiological studies show alcohol is a factor in 27% of assaults.

For more information: