Various media sources report on the worrying extent of the alcohol problem in the UK, with the BBC and the Sun reporting that 1 in 5 people in hospital beds are heavy drinkers.
This follows a review that pooled the results from 124 studies that looked at the rate of alcohol-related conditions among 1.7 million patients in UK hospitals.
The conditions included alcohol dependence, intoxication, alcohol-related mental or behavioural effects, and liver and stomach problems.
Overall, these problems caused by alcohol affected 1 in 5 hospital patients. And 1 in 10 patients were diagnosed as having alcohol dependence.
But the results across individual studies varied widely, and overall the researchers considered this to be very low-quality evidence.
Prevalence tended to be much higher among younger people and in A&E or mental health settings than general hospital wards.
Nor does this mean that 1 in 5 people in the UK now have an alcohol misuse problem.
The rate among a sample of hospital patients is always likely to be higher than among the general population.
Nevertheless, the research may highlight the need for hospital-based health professionals to be aware of those at risk from harmful alcohol use and make sure they're getting the support they need.
The study authors also suggested that inpatient alcohol care teams may be beneficial.
Where did the story come from?
This study was conducted by researchers from King's College London and the Maudsley NHS Foundation Trust, the London School of Hygiene and Tropical Medicine, the University of Brighton and the University of Sussex.
It was supported by a Medical Research Council Addiction Research Clinical Fellowship given to 1 of the authors.
The media coverage is generally representative of the research, highlighting the potential cost of this problem to the NHS.
But the stories may benefit from recognising that these are very uncertain estimates and cannot be applied to the general population.
What kind of research was this?
As the researchers say, alcohol use prevalence among hospital patients is believed to be higher than among the general population, but past studies have been small with mixed findings, and little attempt has been made to pool their findings.
A systematic review is the best way of gathering the evidence to date on a particular topic, but the pooled findings are only as reliable as the studies included.
What did the research involve?
The authors searched literature databases to identify any studies that:
- were conducted in UK hospitals
- clearly defined the hospital setting (general medical or surgery wards, intensive care, A&E or mental health units)
- reported the prevalence of any of 26 alcohol-related conditions among hospital patients
The 26 conditions were coded using a recognised diagnostic system (International Classification of Diseases version 10, ICD 10) and included broad groups of:
- mental and behavioural disorders caused by alcohol (such as intoxication, dependence or withdrawal)
- accidental or intentional "poisoning" caused by alcohol
- liver disorders caused by alcohol (such as fatty liver, cirrhosis or liver failure)
- gastrointestinal disorders caused by alcohol (such as inflammation of the stomach or pancreas)
- other related disorders (such as nerve and muscle problems)
The researchers excluded studies conducted in specific substance use settings, such as rehabilitation or detox units.
What were the basic results?
The researchers identified 124 studies covering 1.7 million patients.
The majority of studies were cross-sectional, just looking at the prevalence of alcohol-related conditions at 1 point in time.
The overall pooled estimate of any alcohol-related condition among hospital patients was 19.8%, or 1 in 5 people. Alcohol dependence specifically affected 10.3%, or 1 in 10.
But prevalence rates across the individual studies were highly variable and the pooled results were considered to be very low-quality evidence.
Most of the variation in results came down to study setting: prevalence was generally higher among A&E and mental health settings (around 30 to 40%) compared with general medical or surgical wards (around 10 to 20%).
Patient age also accounted for the variable results: each 1-year increase in age above the age of 18 was linked with reduced prevalence.
Surprisingly, although the 124 studies spanned over 40 years from the 1970s to date, there was little change in prevalence over the years.
How did the researchers interpret the results?
The researchers concluded: "An estimated 1 in 5 patients in the UK hospital system use alcohol harmfully, and 1 in 10 are alcohol-dependent."
This is a valuable study that has pooled the results of UK studies to date that have reported the prevalence of alcohol-related conditions among UK hospital patients.
With an overall prevalence of 1 in 5, this seems much higher than may have been expected.
The prevalence seems particularly high when you consider that the study only looked at conditions that are the certain result of alcohol misuse (like dependence or alcohol liver diseases) rather than conditions that might be linked to alcohol intake (like certain cancers or high blood pressure).
But these estimates are very uncertain. There was high variation in the prevalence rates among different studies, settings and populations.
The prevalence rates among people in A&E and mental health settings were much higher than on general hospital wards.
Pooling the results has "pulled up" or increased the estimate for the overall hospital population.
We should not assume that 1 in 5 people in any general hospital bed in the UK will have alcohol problems.
Similarly, the 1 in 5 prevalence rate should not be applied to the general UK population.
The prevalence of alcohol-related conditions among any hospital sample is always likely to be higher than the prevalence among any general population sample.
Another point, which the researchers acknowledge, is that few studies reported how they diagnosed these alcohol-related conditions.
So although they used recognised diagnostic codes, there could be some under- or over-reporting in different studies.
And it's worth noting that the media reporting may imply this is an increasing problem.
But whether the study was conducted in the 1970s or the 2010s made little difference to the prevalence.
This appears to be a long-standing problem that's been the same for decades. It's just now we have a review pooling the findings for the first time.
If you're concerned about your drinking, support is available.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
BBC News, 4 July 2019
Guardian, 4 July 2019
Sun, 4 July 2019
Mail Online, 4 July 2019
Links to the science
Addiction. Published online 3 July 2019