Although alcohol screening and counseling of patients who consume higher-than-recommended amounts of alcohol have been shown to reduce drinking, counseling rates are low. Physicians cite many barriers to alcohol counseling, including lack of knowledge, discomfort with the issue, lack of time and training, and other concerns. McCormick and colleagues looked at actual physician behavior recorded by audiotape (rather than self-report) to determine how physicians discussed alcohol consumption with their patients.

This qualitative study was part of the Ambulatory Care Quality Improvement Project (ACQUIP), which was designed to assess the approach to six medical conditions. In this case, alcohol misuse (see accompanying table) was studied at a single site, with audiotapes taken of patients who screened positive for drinking above the recommended limits in the past year. Of the 7,700 patients eligible for the ACQUIP study, 840 (11 percent) completed the baseline survey and screened positive for alcohol misuse. Of these, 68 encounters involving 47 patients were audiotaped, with 39 encounters (29 patients) mentioning alcohol. Although the patients and physicians consented to being audiotaped, neither knew that the study was geared toward alcohol misuse.

The researchers found that alcohol discussions lasted a median of 45 seconds. Although patients often disclosed potentially risky drinking behaviors, prompted and unprompted, physicians were not likely to follow up on these disclosures; they dismissed or minimized the issue. Physicians were likely to focus on nonalcohol-related topics, even when the patient linked these to concerns about drinking. If alcohol-drinking patterns were pursued, it was through closed questioning.

When physicians did give advice, it tended to be vague, without mentioning specific drinking limits. Finally, physician body language, as well as uncomfortable laughter or manner of speaking, suggested discomfort with the subject. Of note, physicians were much more assertive and direct when discussing smoking cessation.

Although patients often disclose information about their drinking, physicians do not adequately counsel them. In this study, physicians did not respond, or if they did, they were vague in their advice and appeared uncomfortable with the subject. The authors conclude that because alcohol screening and counseling has been shown to be beneficial to at-risk patients, efforts should be directed at improving physicians' communication about this issue.