New study suggests clinicians overlook alcohol problems if patients are not intoxicated

Medical staff struggle to spot problem drinking in their patients unless they are already intoxicated, according to research by the University of Leicester.
The work led by Dr Alex J Mitchell, consultant at Leicestershire Partnership NHS Trust and honorary senior lecturer at the University, reveals that clinical staff often overlook alcohol problems in their patients when they do not present intoxicated.
In a new study involving 20,000 patients assessed for alcohol problems by medical staff, all clinicians struggled to detect alcohol problems whether or not patients volunteered information regarding their drinking.
1 in 4 of the adult population in England (33% of men and 16% of women) consumes alcohol in a way that is potentially harmful to their health and 6% of men are alcohol dependent (Pilling et al, 2011). 1 in 6 primary care patients have an alcohol use disorder or are alcohol dependent.
General practitioners (GPs) identified 40% of problem drinkers, hospital doctors identified 50% of problem drinkers and mental health specialists recognised 55% of problem drinkers. Clinicians correctly recorded a diagnosis in the case-notes for only 1 in 3 people who had an alcohol problem. Only alcohol intoxication was accurately identified. A&E clinicians were able to correctly detect patients with alcohol intoxication in 9 out of 10 patients. In research studies where patients admitted to a drinking problem by self report, the same rates of under-detection occurred.
Assessing for alcohol problems in patients using a short questionnaire is recommended by the UK Primary Care Service Framework and NICE but not widely implemented by clinicians.
Dr Alex Mitchell said: ‘This study highlights that clinical identification of alcohol problems is challenging in busy clinical environments. When clinicians try and spot alcohol problems they often miss patients who have serious alcohol problems but who are not currently intoxicated. Further they can misidentify about 5% of