Mobile Health for Alcohol Use Disorders in Clinical Practice

Percentage of self-reported Heavy Drinking Days (HDD: 5 or more standard drinks on any day for men under 65, 4 or more standard drinks for women and men over 65) in the last 7 days

The physical health subscale of Patient Reported Outcomes Measurement Information System (PROMIS) Global-10 short form was used to assess the quality of life-physical health (QoL-PH). It consists of 4 questions that assess overall physical health aspect of quality of life. Each question will ask patients to indicate their physical health on a scale of 1 (poor) to 5 (excellent). The raw scores of physical health scores (range: 0 to 20) are converted to T-scores (16.2 to 67.7). Higher scores mean better physical health aspects of quality of life.

The mental health subscale of Patient Reported Outcomes Measurement Information System (PROMIS) Global-10 short form was used to assess the quality of life-mental health (QoL-MH). It consists of 4 questions that assess overall mental health aspect of quality of life. Each question will ask patients to indicate their mental health on a scale of 1 (poor) to 5 (excellent). The raw scores of mental health scores (range: 0 to 20) are converted to T-scores (21.2 to 67.6). Higher scores mean better mental health aspects of quality of life.

Patients in the clinically-integrated arm will have the option to share their data collected by A-CHESS with the health coach. Knowing how many patients are willing to share data about their drinking will allow researchers to make hypotheses about how effective A-CHESS can be in a healthcare setting. We did not collect data for this metric.

Patients will be asked to indicate hospital stay (nights) in the past 6 months. Researchers will include this in the cost-effectiveness analysis.

The cost (in US dollars) of the intervention will be calculated to determine cost-effectiveness of the study.

The Alcohol Use Disorders Identification Test (AUDIT) screening tool was used to assess patient alcohol severity. The screening tool consists of 10 questions with 5 possible answers (scoring 0-4). The sum scores of the 10 items are the scale scores ranging from 0 to 40. Higher scores indicate greater alcohol severity.

The Diagnostic and Statistical Manual of Mental Disorders 5th edition criteria of alcohol use disorder will be used to assess the patient's alcohol use disorder severity. Each item has a score of 1 for agreement or 0 for disagreement of a statement. The sum score of all 11 items is the scale score (range of 0-11 - higher scores indicate greater severity). The severity of the AUD can also be defined as 1) mild: 2 to 3 symptoms; 2) moderate: 4 to 5 symptoms; 3) severe: 6+ symptoms.

The study will assess how connected patients feel with others. Greater connectedness is associated with fewer risky drinking days. Patients are asked five Likert-scale questions about their relationships with others, responding on a 1-5 scale from 'Never' to 'Nearly Always.' Mean scores are calculated as scale scores, ranging from 1 to 5. Higher scores indicate greater perceived relatedness with others.

The study will assess patients' perceptions of their ability to control and reduce their drinking in order to meet their alcohol use goals. Patients tend to drink less when they feel more confident in managing their alcohol use. Patients will respond to four questions from the Perceived Competence Scale (PCS) about their confidence in responsibly managing alcohol use in various situations. Responses are on a 1-7 scale, from 'Not at all' to 'Very true.' Mean scores (ranging from 1 to 7) are used as the scale scores, with higher scores indicating greater confidence in controlling or reducing drinking.

The study assesses how motivated patients are by their own goals. Patients typically drink less when they, themselves, want to reduce their alcohol. Patients will be asked 6 questions from the autonomous motivation subscale of the Treatment Self-Regulation Questionnaire (TSRQ) about why they will reduce their alcohol use and can indicate how much they identify with each statement on a 1-5 scale corresponding to "Not true" to "Very true", respectively. The mean scores are the scale scores, ranging from 1 to 5. Higher scores mean more autonomous motivation to control and reduce alcohol use.

The time that participants spent using Tula will be assessed.

The patient's ability to meet their weekly goal will be assessed. When patients are not on track to meet their weekly goal a setback alert will be triggered.

The variables that help predict a setback will be assessed and reported. Knowing the variables for setbacks will help researchers program A-CHESS to better predict when patients are likely to have a setback.

Clinicians and patients will be alerted when the system predicts that they may have a setback. This allows researchers to know how useful patients and health coaches found the alerts.

The number of pages viewed by patients and health coaches will be assessed know how useful patients found different content.

Patients will take a revised Brief Alcohol Monitor survey weekly that asks patients to indicate whether any of the five protection factors exist (sleep, mood, social interaction, work, and urges to drink). The protective factor score is the sum of the number of protection factor items reported by patients during the study period. The scores of the protective factors were planned to be used in the lapse prediction model. Since the study plan changes and the lapse prediction was not the focus of the study, only descriptive statistics are reported. No additional statistical analysis was conducted. Scores range from 7-35. Lower scores indicate a worse outcome for each factor.