As explained in a recent post on Public Health England's blog and our initial report, the all-encompassing alcohol-related hospital admissions indicator we came to know and love now has a new slimline companion, to be published in the Public Health Outcomes Framework.
This follows a consultation by PHE last year after the PHOF identified a "preferred option" of an indicator "based on just alcohol-related primary diagnoses".
The new headline indicator is simpler than the previous, although it still relies on 'attributable fractions', which are explained here by the Local Alcohol Alcohol Profiles for England (LAPE) boffins. Attributable fractions are just a scientific way of saying that, on average, some hospital admissions are more related to alcohol than others. This allows calculations as to the overall contribution of alcohol to hospital admissions given its wide health impact.
Hospital admissions are coded for the purposes of data collection according to ICD-10, and each admission is assigned a primary code and up to 19 secondary codes for any diagnoses that affect the treatment needed. The secondary codes may include 'external cause' codes such as motor accident or fall - ie describing the circumstance of the patient that lead to admission.
The original indicator searches every hospital admission for a code that is linked to alcohol and applies the attributable fraction to it. Where an admission has more than one alcohol-related code, the code with the highest attributable fraction is used. The sum of all these fractions of admissions is then used to calculate the rate of alcohol-related admissions for a given population.
The new indicator uses a much narrower search: it looks only for primary or external cause codes that relate to alcohol. These are counted in the same way as before, ie by applying attributable fractions.
Some examples that illustrate the difference between the old and new indicators are highlighted below:
- Patient admitted with a broken leg but also has hypertension (and no alcohol-attributable external cause codes) - IN OLD INDICATOR BUT NOT IN NEW INDICATOR
- Patient admitted with hypertension and various other conditions but hypertension is recorded as the primary code (or reason for admission) - IN OLD INDICATOR, AND IN NEW INDICATOR
- Patient admitted for head injury (primary code) and acute intoxication (secondary code) but no external cause codes - IN OLD INDICATOR BUT NOT IN NEW INDICATOR (comment: an unfortunate anomaly in the new methodology)
- Patient admitted for head injury (primary code) and a car accident (secondary <external cause> code) - IN OLD INDICATOR AND IN NEW INDICATOR
- Patient admitted for acute intoxication (primary code) - IN OLD INDICATOR AND IN NEW INDICATOR
Pros and Cons?
Clearly the new indicator will produce a much smaller headline figure of alcohol-related admissions. As we highlighted last year, this revised annual figure is likely to come in under 300,000 - way down from the previous annual headline figure of well over one million. This could result in far less press attention to the issue of alcohol-related admissions, which may please critics of the previous one.
PHE argue that the new methodology will be more sensitive to local action on alcohol-related harm (by including a higher proportion of acute admissions due to alcohol), and less influenced by changes in coding practice. However, others fear that the impact of excessive drinking on chronic conditions will be sidelined by the new indicator. See the PHE blog Understanding alcohol-related hospital admissions and associated comments for more on this.
Comments