PREDICTORS OF TREATMENT COMPLETION AND DISCONTINUATION IN GROUP TREATMENT OF PATIENTS WITH AUD: RESULTS FROM A NORWEGIAN MULTI-SITE PREDICTOR STUDY

PREDICTORS OF TREATMENT COMPLETION AND DISCONTINUATION IN GROUP TREATMENT OF PATIENTS WITH AUD: RESULTS FROM A NORWEGIAN MULTI-SITE PREDICTOR STUDY
Kristoffer Høiland1, Espen Kristian Ajo Arnevik2, Katie Witkiewitz 3,4 & Jens Egeland1,5
Affiliations: 1Vestfold Hospital Trust, Norway, 2Section for Clinical Addiction Research, Oslo University Hospital, Norway, 3Department of Psychology, University of New Mexico, USA, 4Center on Alcohol, Substance Use and Addictions, University of New Mexico, USA,  and 5Department of Psychology, University of Oslo, Norway
Background: A large percentage of patients with alcohol use disorder (AUD) discontinue treatment. More knowledge of predictors of discontinuation would be beneficial for clinicians in preventing premature discontinuation and tailoring treatment to the patient.
Aim: To investigate predictors of completion and discontinuation in a sample from a multi-site study on outcome predictors of AUD treatment. Most of the baseline data in the study are registered, and we have started analyzing the data. The analyses will be completed in due time before the conference. Preliminary analyses presented briefly in this abstract indicate some significant predictors of discontinuation.
Methods: A total of 136 patients have been recruited in a multi-site study on treatment outcome predictors in group treatment of AUD (Clinical Trials.gov ID NCT04822987). 
Patients were examined at baseline with measures of alcohol use, drug use, psychological symptoms, maladaptive personality functioning, and cognitive functioning. We used validated measures available in clinical practice: AUDIT, DUDIT, SCL-90R, SIPP-118, MoCA, and subtests from WAIS-IV (Block design, Similarities, Digit span, Visual puzzles, and Information) and D-KEFS (The Color word test [CWT] conditions 1-4 and The Trail making test [TMT], conditions 2-4). In addition, demographic factors were registered. These validated measures were examined as predictors of completion versus discontinuation in univariate (ANOVA) and multivariate analyses (logistic regression).
Results: Preliminary analyses show that composite measures of executive functioning (TMT 4, CWT 3 and 4) and mental processing speed (TMT2, 3 and CWT 1, 2) predict discontinuation from treatment, such that lower executive function and reduced mental processing speed predicted discontinuation.
Conclusion: Patients were examined with instruments readily available to most clinicians to investigate predictors of discontinuation. Clinicians should examine executive functioning and speed before the patient commences treatment, as these are predictors of discontinuation. More knowledge about these factors may help detect patients at risk of discontinuing treatment, make prognostic evaluations, and carefully tailor treatment to at-risk patients.