A heavy drinking day is defined as 5 or more drinks for men and 4 or more drinks for women during a 24 hour period.

Intention to Treat

Timeline Follow-back drinking data is used to calculate the % of days abstinent per week during Weeks 3-11

Timeline Follow Back daily drinking data used to calculate the weekly mean drinks per drinking day

Timeline Follow Back daily drinking data used to calculate the weekly mean drinks per day

Timeline Follow Back data used to calculate the % of very heavy drinking days per week. Heavy drinking is 10+ drinks per day for females and 12+ drinks per day for males

Timeline Follow Back data used to calculate the % of subjects that maintained abstinence weeks 3-11.

Timeline Follow Back data used to calculate the % of subjects that didn't have a heavy drinking day during study weeks 3-11.

Drinkers Inventory of Consequences (DrInC) - Alcohol-related problems are determined using the DrInC (Miller et al., 1995). The DrInC is a self-administered 50-item questionnaire designed to measure adverse consequences of alcohol abuse in five areas: Interpersonal, Physical, Social, Impulsive, and Intrapersonal. Each scale provides a lifetime and past 3-month measure of adverse consequences, and scales can be combined to assess total adverse consequences. We used a modified version of the DrInC that just included the 45-items that summed the Interpersonal, Physical, Social, and Impulsivity items. This total score (min=0, max=135) was analyzed in this study with high scores indicative of more alcohol-related consequences (a poor outcome for a given study participant). The DrInC was assessed at study weeks 6 and 12. Analyses averaged across these weeks.

The Penn Alcohol Craving Scale (PACS) is a five-item, self-report measure that includes questions about the frequency, intensity, and duration of craving, the ability to resist drinking, and asks for an overall rating of craving for alcohol for the previous week (Flannery et al., 1999). The summed total score of the 5 items was used in the analysis (min=0, max=30) with higher scores indicative of higher craving for alcohol (a poor outcome). Based on clinical study results, the PACS has been shown to be a reliable and valid measure of alcohol craving and can predict subjects at risk for subsequent relapse. The PACS was assessed at study weeks 4, 6, 8, 10, and 12. Analyses averaged across these weeks.

The MADRS is an observer rating scale that has proven to be an efficient and practical measure of depression (Montgomery and Asberg, 1979). The scale was constructed to be sensitive to changes in treatment effects. Its capacity to differentiate between responders and non-responders to antidepressant treatment has been shown to be comparable to the Hamilton Rating Scale for Depression, another established measure of depressive symptomatology, but the MADRS has greater sensitivity to change during the course of a depressive phase. It has exhibited high inter-rater reliability and appears to be oriented more towards psychic as opposed to somatic aspects of depression. The MADRS is the sum of the 10-item in a checklist where items are rated on a scale of 0 to 6 with anchors at 2-point intervals. Scores range from 0 to 60. Higher scores are indicative of greater depressive symptoms (a poor outcome). The MADRS was assessed at weeks 4, 6, 8, 10, and 12. Analyses averaged across these weeks.

The Hamilton Anxiety Scale consists of 14 items, each defined by a series of symptoms. Similar to the HAM-D, each item is rated on a 5-point scale, ranging from 0 (not present) to 4 (severe) (Guy, 1976). A total score is derived from the summed items (min=0, max=56) with higher scores indicative of greater anxiety (a poor outcome). The HAM-A was assessed at study weeks 4, 6, 8, 10, and 12. Analyses averaged across these weeks.

The PSQI is a 19-item questionnaire assessing the subject's overall sleep experience in the past 30 days (Buysse et al-1989). The lower the overall score, the better the person sleeps. The tool has an adequate internal reliability, validity and consistency for clinical and community samples of the various populations. Range is (0-21); >6 indicative of "poor" sleep quality. The PSQI was assessed at study weeks 4,8, and 12. Analyses averaged across these weeks.

The SF-12 will be used to assess overall health status. The SF-12 is a 12-item questionnaire developed in 1994 as a shorter alternative to the SF-36 to reproduce the physical and mental health summary measures with at least 90% accuracy. We calculated the physical and mental component summary scores which were both converted to T-scores (min=0, max=100) normed to the general population such that a T=50 is the average score in the general population. Higher scores are indicative of better health status.

The SF-12 will be used to assess overall health status. The SF-12 is a 12-item questionnaire developed in 1994 as a shorter alternative to the SF-36 to reproduce the physical and mental health summary measures with at least 90% accuracy. We calculated the physical and mental component summary scores which were both converted to T-scores (min=0, max=100) normed to the general population such that a T=50 is the average score in the general population. Higher scores are indicative of better health status.

mITT