Doctor, Counselor, Cost-Cutter

  • Primary-care doctors do not typically talk to their patients about problem drinking
  • A new study tests the effectiveness of doctor-initiated advice generated by a routine patient visit
  • Advised patients show a significant decrease in alcohol use, accidents, and health-care utilization
  • Benefit-cost analysis estimates a $43,000 reduction in future health-care costs for every $10,000 invested in early intervention

People who drink above what could be considered a healthy level - that is, more than two to three drinks per day - are at risk for a number of health and safety problems. Excessive alcohol use has been implicated as a cause of liver disease, stroke, cancer, infant neurodevelopmental disorders, and hospital admissions in older adults as well as a leading factor in domestic violence, marital conflict, child abuse, accidents and injuries. A study in the January issue of Alcoholism: Clinical & Experimental Research takes a method also used for smoking cessation - physician intervention - and applies it to problem drinking.

"I was interested in finding out if what I did as a physician made a difference with my patients," said Michael F. Fleming, director of the Family Medicine Research Program at University of Wisconsin - Madison and lead author of the study. "What happens when I talk to my patients for a few minutes about their drinking? Do they decrease their alcohol use? Do they have fewer health problems? Are they hospitalized less often? Do they get into fewer accidents?"

During a routine visit to the doctor, Wisconsin patients (ages 18 - 65) were given a questionnaire to establish at-risk alcohol behaviors. Of the 774 who screened positive, 382 were assigned to a control group and 392 received an intervention program called Project TrEAT (Trial for Early Alcohol Treatment), a protocol originally developed in England that was modified by Fleming and his co-authors. Project TrEAT consists of two 15-minute face-to-face physician conversations, followed by two five-minute nurse phone calls. Project components include a review of ‘acceptable’ drinking, patient-specific alcohol effects, a worksheet on drinking cues, cards to record drinking habits, and a drinking agreement. Forty-eight months later, researchers examined the success and performed a benefit-cost analysis of the project.

"If physicians spend five to 10 minutes talking to their patients about alcohol use," said Fleming, "15 to 20 percent of their patients will significantly decrease their alcohol use, health care utilization, risk of accidents and overall health care costs. Physicians who spend a few minutes talking to their patients about their alcohol problems can make a difference. I would like to see physicians regularly ask all of their patients with mental health, medical and family problems how much they drink - especially if they are going to prescribe medication, because many medications interact directly with alcohol."

"Research on the impact of physician advice about alcohol use," said Jeffrey H. Samet, associate professor of medicine and public health at Boston University," has built upon the knowledge base that problem drinking is common among patients going to see their regular physician, that identification of these drinkers can be accomplished by a few key brief questions, and that physicians do not as yet regularly incorporate national recommendations for screening and brief intervention for alcohol problems."

Samet added that most physicians likely ask new patients about alcohol use but, even so, only a minority of physicians use the best tools available to carry out this task. "Formal screening tests generally require less than one minute to administer," he said. Furthermore, he added, offering advice to problem drinkers is not routinely provided.

"Reasons as to why these medical activities have not as yet been fully embraced," said Samet, " include physicians' lack of confidence in alcohol history taking, lack of familiarity with expert guidelines, concern that patients will object, and the typical lack of reimbursement for this physician activity. Changing physician behavior is not an easy task but not an impossible one either. The time has come for health systems to prioritize implementation of screening for alcohol problems in the primary care setting and delivering brief interventions to those that can benefit. This will require broad training of physicians, particularly those in primary care specialties, in order to ask about and address patients' alcohol problems. This training should occur in medical schools, residency-training programs, and in clinical practice. Finally, physician reimbursement for this work must occur."

Costs of physician reimbursement could be offset by projected savings in future systemic costs. The study estimates a $43,000 reduction in future health-care costs for every $10,000 invested in early intervention. "These benefits could still be appreciated four years after the intervention," noted Samet. "This is an impressive cost savings from a societal perspective. Furthermore, this kind of cost savings for a medical intervention is very uncommon in medicine."

"We know there are 30 to 40 million Americans who drink too much," added Fleming, "with 100,000 of these Americans dying each year because of their drinking. There is also the effect these persons have on their families, their co-workers and on innocent persons killed on our streets and highways. If physicians would conduct brief intervention with these individuals, we could expect a significant reduction in alcohol-related harm in the United States."

Funding for this Addiction Science Made Easy project is provided by the Addiction Technology Transfer Center National Office, under the cooperative agreement from the Center for Substance Abuse Treatment of SAMHSA.

Articles were written based on the following published research:

Fleming, M.F., Mundt, M.P., French, M.T., Manwell, L.B., Stauffacher, E.A., & Barry, K.L. (2002, January). Brief physician advice for problem drinkers: Long-term efficacy and benefit-cost analysis. Alcoholism: Clinical and Experimental Research, 26(1), 36-43.