Validation of an Online Version of the Alcohol Use Disorders Identification Test (AUDIT) for Alcohol Screening in Spanish University Students

Online alcohol screening may be helpful in preventing alcohol use disorders. We assessed psychometric properties of an online version of the Alcohol Use Disorders Identification Test (AUDIT) among Spanish university students. We used a longitudinal online survey (the UNIVERSAL project) of first-year students (18–24 years old) in five universities, including the AUDIT, as part of the WHO World Mental Health International College Student (WMH-ICS) initiative. A reappraisal interview was carried out with the Timeline Followback (TLFB) for alcohol consumption categories and the Mini International Neuropsychiatric Interview (MINI) for alcohol use disorder. Reliability, construct validity and diagnostic accuracy were assessed. Results: 287 students (75% women) completed the MINI, of whom 242 also completed the TLFB. AUDIT’s Cronbach’s alpha was 0.82. The confirmatory factor analysis for the one-factor solution of the AUDIT showed a good fit to the data. Significant AUDIT score differences were observed by TLFB categories and by MINI disorders. Areas under the curve (AUC) were very large for dependence (AUC = 0.96) and adequate for consumption categories (AUC > 0.7). AUDIT cut-off points of 6/8 (women/men) for moderate-risk drinking and 13 for alcohol dependence showed sensitivity/specificity of 76.2%/78.9% and 56%/97.5%, respectively. The online version of the AUDIT is useful for detecting alcohol consumption categories and alcohol dependence in Spanish university students.

Python Package abstcal: An Open-Source Tool for Calculating Abstinence from Timeline Followback Data

In smoking cessation clinical trials, timeline followback (TLFB) interviews are widely used to track daily cigarette consumption. However, there are no standard tools for calculating abstinence based on TLFB data. Individual research groups have to develop their own calculation tools, which is not only time- and resource-consuming but might also lead to variability in the data processing and calculation procedures. To address these issues, we developed a novel open-source Python package named abstcal to calculate abstinence using TLFB data. This package provides data verification, duplicate and outlier detection, missing-data imputation, integration of biochemical verification data, and calculation of a variety of definitions of abstinence, including continuous, point-prevalence, and prolonged abstinence. We verified the accuracy of the calculator using data derived from a clinical smoking cessation study. To improve the package’s accessibility, we have made it available as a free web app. The abstcal package is a reliable abstinence calculator with open-source access, providing a shared validated online tool to the addiction research field. We expect that this open-source abstinence calculation tool will improve the rigor and reproducibility of smoking and addiction research by standardizing TLFB-based abstinence calculation.

The K Nearest Neighbor Algorithm for Imputation of Missing Longitudinal Prenatal Alcohol Data

Background — Missing data are a source of bias in epidemiologic studies. This is problematic in alcohol research where data missingness is linked to drinking behavior. Methods — The Safe Passage study was a prospective investigation of prenatal drinking and fetal/infant outcomes (n=11,083). Daily alcohol consumption for last reported drinking day and 30 days prior was recorded using Timeline Followback method. Of 3.2 million person-days, data were missing for 0.36 million. We imputed missing data using a machine learning algorithm; “K Nearest Neighbor” (K-NN). K-NN imputes missing values for a participant using data of participants closest to it. Imputed values were weighted for the distances from nearest neighbors and matched for day of week. Validation was done on randomly deleted data for 5-15 consecutive days. Results — Data from 5 nearest neighbors and segments of 55 days provided imputed values with least imputation error. After deleting data segments from with no missing days first trimester, there was no difference between actual and predicted values for 64% of deleted segments. For 31% of the segments, imputed data were within +/-1 drink/day of the actual. Conclusions — K-NN can be used to impute missing data in longitudinal studies of alcohol use during pregnancy with high accuracy.

The K Nearest Neighbor Algorithm for Imputation of Missing Longitudinal Prenatal Alcohol Data

Background — Missing data are a source of bias in epidemiologic studies. This is problematic in alcohol research where data missingness is linked to drinking behavior. Methods — The Safe Passage study was a prospective investigation of prenatal drinking and fetal/infant outcomes (n=11,083). Daily alcohol consumption for last reported drinking day and 30 days prior was recorded using Timeline Followback method. Of 3.2 million person-days, data were missing for 0.36 million. We imputed missing data using a machine learning algorithm; “K Nearest Neighbor” (K-NN). K-NN imputes missing values for a participant using data of participants closest to it. Imputed values were weighted for the distances from nearest neighbors and matched for day of week. Validation was done on randomly deleted data for 5-15 consecutive days. Results — Data from 5 nearest neighbors and segments of 55 days provided imputed values with least imputation error. After deleting data segments from with no missing days first trimester, there was no difference between actual and predicted values for 64% of deleted segments. For 31% of the segments, imputed data were within +/-1 drink/day of the actual. Conclusions — K-NN can be used to impute missing data in longitudinal studies of alcohol use during pregnancy with high accuracy.

HealthCall: Smartphone Enhancement of Brief Interventions to Reduce Heavy Drinking In HIV Care

BackgroundHeavy drinking among people living with HIV (PLWH) worsens their health outcomes and disrupts their continuum of care. Brief interventions to reduce heavy drinking in primary care are effective, but in heavy-drinking PLWH, more extensive intervention may be needed. Lengthy interventions are not feasible in most HIV primary care settings, and patients seldom follow referrals to outside treatment. Utilizing visual and video features of smartphone technology, we developed the “HealthCall” app to provide continued engagement after brief intervention, in order to reduce drinking and improve other aspects of HIV care while making minimal demands on providers.MethodsAlcohol-dependent patients at a large urban HIV clinic were randomized to one of three groups: (1) Motivational Interviewing (MI) plus HealthCall (n=39), (2) NIAAA Clinician’s Guide (CG) plus HealthCall (n=38), or (3) CG-only (n=37). Baseline drinking-reduction interventions were ∼25 minutes, with brief (10-15 min) check-in sessions at 30 and 60 days. HealthCall involved daily use of the smartphone for 3-5 min/day, covering drinking and other aspects of the prior 24 hours. Outcomes assessed at 30 and 60 days, and 3, 6 and 12 months, included drinks per drinking day, drinks per day, and days drank, using the Timeline Followback. Analysis were conducted using generalized linear mixed models with pre-planned contrasts.ResultsStudy retention was excellent (85%-94% across timepoints) and unrelated to treatment arm or patient characteristics. During treatment, patients in MI+HealthCall drank less than others (p=0.07-0.003). However, at 6 and 12 months, drinking was lowest among patients who had been in CG+HealthCall (p=0.04-0.06).ConclusionDuring treatment, patients in MI+HealthCall drank less than patients in the CG conditions. However, at 6 and 12 months, drinking was lower among patients in CG+HealthCall. Given the importance of drinking reduction and the low costs and time required for HealthCall, pairing HealthCall with brief interventions within HIV clinics merits widespread consideration.

Baseline Drinking Patterns in Non-Treatment Seeking Problem Drinkers

Aims Natural processes of change have been documented in treatment-seekers who begin to reduce their drinking in anticipation of treatment. The study examined whether non-treatment-seeking problem drinkers would engage in drinking reduction in anticipation of participating in a research study. Methods Non-treatment-seeking problem drinkers (n = 935) were culled from five behavioral pharmacology studies. Participants reported on their alcohol use during the past 30 days using the Timeline Followback. Cluster analysis identified distinct groups/clusters based on drinking patterns over the 30-day pre-visit period. The identified clusters were compared on demographic and clinical measures. Results Three distinct clusters were identified (a) heavy-decreasing drinking group (n = 255, 27.27%); (b) a moderate-stable drinking group (n = 353, 37.75%) and (c) low-stable drinking group (n = 327, 34.97%). The three clusters differed significantly on a host of measures including pre-visit drinking (age at first drink, drinking days, drinks per week, drinks per drinking day), alcohol use severity, alcohol craving, readiness for change, depression and anxiety levels. These differences were alcohol dose-dependent such that the heavier drinking group reported the highest levels on all constructs, followed by the moderate group, and the low drinking group last. Conclusions Baseline drinking patterns of non-treatment-seekers were generally stable and pre-visit reductions were only observed among the heavy drinking group. This generally stable pattern stands in contrast to previous reports for treatment-seeking samples. Nevertheless, the heavier drinking group, which is most similar to treatment-seekers, displayed pre-study drinking reduction. Overall, naturalistic processes of change may pose less of a threat to randomization and testing in this population.

Ping-Pong Housing: Women’s Post-Incarceration Trajectories

This study examines the post-incarceration housing experiences of 33 women. Using Residential Timeline Followback methodology, participants were asked to report where they lived at arrest and every location since their release. Follow-up questions asked women to describe these locations, who they lived with, how much they paid, and whether or not they felt safe. Demographic information and criminal justice history were recorded. The data paint a complicated picture of social and community resources, persistence, and struggle. Housing assets lost at incarceration were difficult to recover. Most women bounced between various locations, relying heavily on short-term subsidized congregate housing programs and rarely securing independent housing. Participants described the family, friends, and acquaintances who housed them during reentry as overextended and vulnerable. Implications for policy and practice are explored.