"It may be argued that subjects who remain abstinent [from alcohol] during treatment are the most successful because psychosocial functioning and physical health depend on sobriety," say study authors Sue-Jane Wang, Ph.D., and Celia Winchell, M.D., and colleagues from the U.S. Food and Drug Administration in Rockville, Md. "However, other patterns of drinking are very common during treatment and many analytic approaches fail to make a distinction among the patterns."
For example, traditional research methods often don't distinguish between a person who drinks only once a week during a 12-week trial and one who abstains for the first 10 weeks but spends the last two weeks intoxicated, according to the study. "These summary measures fail to capture the full complexity of the drinking pattern over time," the researchers say.
They add, "There has been a great deal of contention on whether the effects of alcoholism treatment should be evaluated solely against the criterion of abstinence. … The clinical community is still searching for a better description of what constitutes effectiveness in alcoholism treatment trials. More informative statistical analysis methods are necessary to arrive at meaningful evidence."
Wang and colleagues tested a research tool called the multiple failure time approach that asks more nuanced questions than traditional approaches, including: "Does the treatment reduce the rate of relapse to heavy drinking?" They used this approach to re-examine a study that found weakly statistically significant evidencethat an alcoholism treatment drug called naltrexone was effective.
By taking into account both the time and the frequency of the study participants' drinking episodes, the researchers noted two things that were overlooked in the first analysis: The risk of having any drinking episodes and any heavy drinking episodes was significantly lower in the group treated with naltrexone rather than a placebo.
The study results are published in the December issue of the journal Alcoholism: Clinical and Experimental Research.
FOR MORE INFORMATION
Health Behavior News Service: 202-387-2829 or www.hbns.org.
Interviews: Contact Sue-Jane Wang at wangs@cder.fda.gov.
Alcoholism: Clinical & Experimental Research: Contact Mary Newcomb at 317-278-4765 or mnewcomb@iupui.edu, or visit www.alcoholism-cer.com.
BY ANN QUIGLEY, CONTRIBUTING WRITER
HEALTH BEHAVIOR NEWS SERVICE