Monday, January 11, 2010

Risky Drinking

The one-question questionnaire.

Answer: Once or twice.

Question: “How many times in the past year have you had 5 or more drinks (for men), or 4 or more drinks (for women) in a single day?"

A recent study published in the Journal of Internal Medicine strongly suggests that this simple question identifies those drinkers at risk for alcohol use disorders roughly 75 % of the time.  While the one-question screening test has been endorsed by both the National Institute on Drug Abuse (NIDA) and the National Institute on Alcohol Abuse and Alcoholism (NIAAA) for several years, little clinical evidence existed for assuming that it worked. 

It seems weirdly unrealistic as a standard measure, leaving out, as it does, so many other telling features of active alcoholism. Nonetheless, the group at the Boston University School of Medicine that conducted the research concluded that “the single screening question recommended by the NIAAA accurately identified unhealthy alcohol use in the sample of primary care patients.”

It seems safe to assume that the majority of people who occasionally overdrink are not alcoholics. Is the occasional binge or bender by recreational drinkers really that rare? In a research summary comment on the results, Dr. Peter D. Friedmann opens up the possibility of using the one-question screen to “facilitate more discussion of heavy episodic (binge) drinking, a major source of adverse consequences among nondependent drinkers.”

Five or more drinks, on a single occasion in the past year? Is that really sufficient data? Is a response of  >1 really a genuine cause for concern?

But it gets even stranger. In 2006, the Journal of Studies on Alcohol published a primary care validation study which showed that narrowing the criteria to one incident in three months did not significantly change the results. When the choices available were “within 3 months,” “within 12 months,” “ever,” or “never,” the 3 month and 12 month positive answers were predictive of risky consumption levels about 75% of the time in a study of 625 patients.

One conclusion to be considered is that “normal,” non-alcoholic drinking males rarely—if ever—consume more than 5 drinks in one evening (4 for women). For heavier drinkers, this seems an impossibly Puritan standard, and useless as a diagnostic tool.  As usual, more studies are needed. But the authors of the 2006 paper were confident enough to conclude: “A single question about the last episode of heavy drinking is a sensitive, time-efficient screening instrument that shows promise for increasing alcohol screening in primary care practices.”


red dave said...

From the linked to paper
False positive rate (α) = 1 − specificity
False negative rate (β) = 1 − sensitivity

So with 100 people 5 are alcos. and with this test you will find 27 to be alcos and miss nearly one alco. 27 out of 100 will be classed as alcoholics?

Dirk Hanson said...

"So with 100 people 5 are alcos. and with this test you will find 27 to be alcos and miss nearly one alco. 27 out of 100 will be classed as alcoholics? "
In my opinion, the BBC Alcohol Concern number is probably low, and the figures that emerge from the one-question screen are probably high. My best guess is that the number of alcoholics falls somewhere between the former's 5% and the latter's 27%.

OMDS said...

5 or more drinks on a single occasion could be risky and it effects the normal health system. It's not preferable in any way.

Related Posts Plugin for WordPress, Blogger...