Spaced Versus Massed Alcohol Avoidance Task

The Effectiveness of Spaced Versus Massed Alcohol Avoidance Training in Inpatient Alcohol Use Disorder Treatment: A Multisite Randomized Trial

Radboud University Medical Center, Nijmegen Institute for Scientist-Practitioners in Addiction (NISPA)

Rationale: Alcohol-Avoidance Training (AAT) has been used successfully to strengthen avoidance-tendencies in patients with alcohol use disorder (AUD). AAT is already recommended in German clinical treatment guidelines as an evidence-based treatment for AUD and may be incorporated in the next revision of the Dutch clinical guidelines on AUD. Studies in other fields (semantic learning) suggest that spaced learning may be superior to massed learning, but this has not been studied with regard to AAT. Objectives: To compare the effectiveness of spaced versus massed AAT sessions. Study design: a two armed, randomised controlled trail. All participants will receive AAT in addition to routine clinical care in an inpatient setting (Treatment As Usual; TAU). AAT sessions in the experimental group (AAT-S) will be spaced out over four weeks following detoxification. Sessions in the control group (AAT-M) are massed within one week following detoxification. Assessments of alcohol consumption and craving take place before the start of AAT (baseline: T0, timeframe: last 30 days before admission (alcohol use) or past week (craving)) at three (T1) and six months follow-up (T2). Study population: 200 patients with a primary DSM-5 diagnosis of AUD who receive TAU at three addiction care sites (clinical facilities 'Zevenaar', 'Tiel' and 'Wolfheze') of IrisZorg. Patients have finished alcohol detoxification, age ≥ 18, have good Dutch proficiency and have given written informed consent. During the follow-up assessments they are likely to have progressed to regular outpatient addiction treatment. Intervention: During their four week (minimum) admission all participants receive TAU, which includes Community Reinforcement Approach (CRA) (Meyers & Smith, 1995) grouptraining, AAT, sociotherapy and pharmacotherapy. AAT is a Cognitive Bias Modification paradigm that is used to retrain alcohol approach biases (Eberl et al., 2014). In AAT participants must react to pictures of alcoholic and non-alcoholic beverages with a joystick to the tilt of the pictures which are presented on a computer screen. In current routine clinical care AAT sessions are massed in the first week after detoxification (control condition: AAT-M). In the experimental condition AAT trails will be spaced out over four weekly sessions instead of one week (AAT-S). Main study parameters/endpoints: Changes from baseline to three and six month follow-up in: Mean daily units of alcohol consumed (past 30 days); at baseline this refers to the 30 days directly pre-admission). Mean ratings of mean alcohol craving (past seven days). Nature and extent of the burden and risks associated with participation, benefit and group relatedness: Participants in both conditions will receive the same amount of AAT sessions and trials. Sessions in the AAT-S condition will be spread out over four weeks. Participants in this condition will therefore be exposed to AAT alcohol pictures over a longer period. Given our current experience of AAT as part of TAU, we expect little risk of participants experiencing more sensations of craving. Alcohol (use) is discussed daily during admission as part of TAU. As an extra burden, participants will be asked to complete a questionnaire before the first AAT session. Participants will be approached for follow-up assessment three and six months following the first month of inpatient treatment. Participants receive an incentive (a voucher worth €15,-) after completing all FU assessments, as a compensation for the extra burden.

INTRODUCTION AND RATIONALE 1.1. Dual Process Models in Addiction Dual-process models as described by Satpute and Lieberman (2006), Schneider and Chein (2003), Sherman et al. (2008), Smith and DeCoster (2000) and Strack and Deutsch, (2004) have provided more insight as to how individuals with an addiction can be 'drawn in' by alcohol-related cues at an implicit level, while at the same time more controlled, conscious cognitive processes may be invested in actively trying to avoid substances. Even when patients with substance abuse are made aware of long-term negative consequences, learn to seek alternatives, deal with group-pressure or craving, they are still susceptible to their immediate attention being subconsciously hijacked by substance related cues that trigger an automatic implicit approach-tendency towards the substance (Stacy & Wiers, 2010; Wiers, Van Woerden, Smulders & de Jong, 2002; Wiers et al., 2007; Wiers, Rinck, Dictus, & van den Wildenberg, 2009; Wiers, Gladwin, Hofmann, Salemink, & Ridderinkhof, 2013). Therapeutic interventions that aim to reduce these implicit approach tendencies have been developed. 1.2. Alcohol Avoidance Training Alcohol Avoidance Training (AAT), is one of several Cognitive Bias Modification paradigms that is used to retrain alcohol approach biases (Eberl et al., 2014). Pictures of alcoholic and non-alcoholic beverages are used as stimulus material. Participants are instructed to react with a joystick to the tilt of pictures, presented on a computer screen. For instance, participants make an approach movement (pull) with left-tilted pictures, and make an avoidance movement (push) with right-tilted pictures. Pulling increases the size of the pictures whereas pushing decreases their size. After the correct movement is made, the pictures disappear, otherwise they do not, until the correct response is made. It has been demonstrated in lab studies, that a change in approach bias (increased avoidance of the targeted substance) predicts consumption behaviour of the targeted substance (Stacy & Wiers, 2010). Therefore, when a strong training effect can be demonstrated, a clinical effect is likely to ensue. Research thus far indeed suggests that AAT provides a clinical effective add-on to TAU in the treatment of Alcohol Use Disorder (AUD). The first study on the effectiveness of AAT in problematic alcohol use was conducted in a non-clinical sample of 42 male students, classified as hazardous drinkers (Wiers, Rinck, Kordts, Houben, & Strack, 2010). The results showed that the automatic action tendencies to approach alcohol were reduced. In 2010, the first Randomized Controlled Trial (RCT) using AAT in patients with AUD was conducted. The sample consisted of 214 inpatients. When AAT was added to TAU, the total number of relapses after one year decreased by about 10% compared with patients who had received a sham or no training (Wiers, Eberl, Rinck, Beckers, & Lindenmeyer, 2011). Importantly, the clinical effect was replicated. In addition, analyses revealed four to six AAT sessions to be the mean optimum number of sessions needed (Eberl et al., 2013; Manning et al., 2016). Although the effect sizes found are small to medium (Kakoschke, Kemps & Tiggeman, 2017), the results are interesting given the high relapse rates (Brandon, Vidrine, & Litvin, 2007) in addiction treatment and minor burden related to the intervention. Apart from the replicated clinical effectiveness, in our own clinical experience, AAT can be easily integrated with (inpatient) TAU. It does not require highly trained therapists as is the case in Cognitive Behavioral Therapy (CBT) or Community Reinforcement Approach (CRA). In addition, six AAT sessions will take the participant an average of 60 to 90 minutes total to complete. Therefore, the burden in time is low which may increase engagement and attendance and decrease drop-out. Finally, AAT can be provided as early as during detoxification (Manning, Mroz, Garfield, Staiger, Hall, Lubman, & Verdejo-Garcia, 2019; Manning et al., 2016) and may even be more effective than when provided after detoxification. In comparison, traditional CBT and CRA interventions are thought to be less effective during detoxification as a result of dampened cognitive functioning in attention and often executive functioning as a result of prolonged abuse (Aharonovich, Shmulewitz, Wall, Grant, & Hasin, 2017; Domínguez-Salas, Díaz-Batanero, Lozano-Rojas, & Verdejo-García, 2016). AAT is already recommended in German clinical treatment guidelines and may be incorporated in the next revision of the Dutch clinical guidelines on AUD (the current, outdated version is of 2009). Therefore, AAT has been implemented as part of routine clinical care for AUD treatment in inpatient settings of 'IrisZorg' (a Dutch addiction care provider). 1.3. This study Although the optimum length of alcohol avoidance training has been studied, little is known about the impact of the distribution of these sessions over time. As a result no guidelines or recommendations exist on whether AAT sessions should be spaced out over time or provided in a massed format. In clinical practice (within our inpatient care facilities), AAT as standard treatment is commonly administered during the first week after detoxification. Although, in semantic learning, 'cramming' of learning episodes within a short period of time (known as 'massed practice') has been found to be effective for short term retention. When long term retention is preferable, learning episodes should be spaced out over longer periods of time (Cepeda, Rohrer, Wixted & Pashler, 2008). Optimum criterion performance is achieved when lag between sessions is approximately 10-20% of the desired retention interval. Spacing effects have been demonstrated on numerous explicit measures of memory (Donovan & Radosevich, 1999; Janiszewski, Noel & Sawyer, 2006; Cepeda, Rohrer, Wixted & Pashler, 2006) and sometimes implicit learning (Greene, 1989; Jacoby & Dallas, 1981). Greene (1989) added that spacing effects in implicit leaning are eliminated by incidental learning, thus disproving that the spacing effects apply to all forms of learning. To our knowledge, the impact of spaced learning on the procedural learning that is involved in AAT has never been evaluated in a controlled study. It is not self-evident that distributed practice will be more effective than massed practice. AAT research has thus far shown that even small effect sizes can accumulate to substantial changes in clinical effects. Knowledge about the optimum distribution of AAT session are of importance for the ongoing development of AAT. This study will be used to evaluate the possible superior effect of spaced practice AAT in multiple inpatient addiction care facilities. Effectiveness will refer to the possible impact of spaced practice (AAT-S) compared to the massed practice control group (AAT-M) on clinical outcomes in terms of alcohol consumption and craving. OBJECTIVES Primary objective: To evaluate the effectiveness of AAT-S compared to AAT-M in inpatient alcohol use disorder treatment in terms of changes in: 1.1 Alcohol consumption. Secondary objectives: To evaluate the effectiveness of AAT-S compared to AAT-M in inpatient alcohol use disorder treatment in terms of changes in: 2.1 Alcohol craving. 2.2 Attentional bias or training effect (manipulation check). STUDY DESIGN The study is a RCT with two arms, with a baseline, three and six month follow up assessment, to evaluate the effects of spaced vs massed AAT sessions as part of inpatient, routine clinical care for patients with AUD. For the primary objective (to determine superiority of spaced practice AAT over massed AAT), a sample size of N = 200 will be required. STUDY POPULATION 4.1 Population (base) The study population consists of adult patients (both male and female) with a primary diagnosis of AUD, who receive inpatient care at the clinical facilities of IrisZorg (at 'Zevenaar', 'Tiel' and 'Wolfheze', the Netherlands). Since this group of patients constitutes the largest group in Dutch addiction care facilities, the study not only focuses on a very relevant sample, but also ensures that enrolling enough participants will be a realistic endeavour. The clinics have a combined total of 81 beds for patients with severe, substance use disorders such as alcohol, opiates (heroin, morphine etc.), stimulants (cocaine and amphetamines), GHB and cannabis. In 2019 a total of 761 unique clients were admitted (73% men). Of these patients, 303 (40%) presented themselves with a primary AUD. 4.2 Inclusion criteria In order to be eligible to participate, a participant must meet the following criteria: ● A primary diagnosis of alcohol use disorder (meeting the DSM-5 criteria (American Psychiatric Association, 2013); ● Age of at least 18 years or older; Good Dutch language proficiency; Written informed consent. 4.3 Exclusion criteria Participants who meet any of the following criteria will be excluded from participation: • Severe, current psychiatric symptoms (especially manic, psychotic, suicidal and aggressive symptoms) that may endanger participants or others and jeopardize study adherence (determined on a case-by-case basis. The staff member who will provide the AAT information, informed consent and instruction will always consult one of the principal investigators in that case). Whether participants meet DSM-5 diagnoses is determined by a staff member (a nurse of physician) during the regular admission procedure, using structured interviews. When participants are screened for in- and exclusion criteria (during detoxification), these findings will be taken into account. 4.4 Sample size calculation Several earlier studies have shown that the average effect size of AAT on relapse rates in clinical samples is small to medium, Cohen's d=.19 - d=.31 (Eberl et al., 2014). This translates into an approximate 10% overall reduction at one-year follow-up (FU), while the effect on the implicit alcohol approach bias (AAT training effect) is usually medium-sized. Note that Eberl et al. (2014) used a dichotomous outcome measurement (relapse vs no relapse) when describing these overall reduction rates. Effect sizes on a continuous outcome scale (e.g. average number of drinks in a specified period) might very well be higher. No evidence that we are aware of regarding the effect of spaced versus massed AAT exists. We therefore assume a small effect size of d =.20. Changes in alcohol consumption is the primary outcome parameter. A power of 80% and an alpha of .05 were used when calculating the required sample size using G*Power (version 3.1.9.3; Faul, Erdfelder, Lang & Buchner, 2007), yielding 164 participants for the per-protocol analysis. In anticipation of approximately 30% study dropout, a total of 234 participants are included. TREATMENT OF SUBJECTS Participant receive either AAT-M of AAT-S within a broader context of routine clinical care. AAT-M is an element of current routine care. The only difference in the AAT-S condition is that the AAT sessions are spread out over four weeks. After baseline assessment, AAT-M participants will receive 4 AAT sessions (each with 300 trials) within the space of 8 days (which totals 1200 trails). AAT-S participants will also receive 4 AAT sessions totalling 1200 trails but spaced out over four weeks (one AAT session per week). Treatment decisions in the context of TAU have priority and may override considerations from the research perspective (e.g. if a participant needs to be transferred, this might interfere with assessments or AAT). Staff members explain AAT (to those who will receive it) as a novel add-on intervention as part of routine care, which may help reduce the risk to relapse. Assessments will make use of questionnaires that are also used for Routine Outcome Monitoring (ROM) and standard inpatient monitoring as part of routine clinical care. METHODS 6.1 Study parameters/endpoints 6.1.1 Main study parameter/endpoint To evaluate the effectiveness of spaced AAT compared to massed AAT in inpatient AUD treatment in terms of a possible reduction in alcohol consumption. o Changes in mean units of alcohol consumed in the past 30 days (MATE/MATE - Outcomes (Schippers, Broekman & Buchholz, 2007) section 1: number of regular drinking days x average number of drinks regular drinking day + number of heavy drinking days x average number of drinks heavy drinking day) (baseline vs. 6 month follow-up). Note that the baseline assessment will focus on alcohol use in the last 30 days preceding admission. 6.1.2 Secondary study parameters To evaluate the effectiveness of spaced AAT compared to massed AAT in inpatient AUD treatment in terms of a possible increase in time-to-relapse, and decrease in craving and alcohol approach bias. o Duration of abstinence - estimated time-to-relapse following discharge (weeks), assessed at three month and six month follow-up. o Changes in mean craving score over the past 7 days (MATE/MATE - Outcomes section 'Q1' from the OCDS-5 (Schippers et. al, 1997): thoughts and desires concerning wanting to use alcohol (baseline vs. three and six month follow-up). o Changes in alcohol approach bias (training effect) over the course of the four AAT sessions. 6.1.3 Other study parameters Other independent variables will be collected to describe the population and may be used to explore potential predictors.: o Age o Years problematic drinking (MATE section 1) o DSM-5 diagnosis (number of comorbid disorders) o Use of anti-craving medication during inpatient, AAT period (acamprosate, naltrexone, baclofen, topiramate, gabapentin) Use of alcohol abstinence enforcing medication during inpatient, AAT period (antabuse, refusal). o Gender o Social-economic status o Marital status o Highest educational level attained o Ethnicity o Substance use (other than alcohol) last 30 days before detox (MATE section 1) o Years substance use (other than alcohol)(MATE section 1) o Time in treatment (total and last inpatient period) (patient files) o Addiction treatment history (MATE section 3) o Facility (Zevenaar, Tiel, Wolfheze) o Length of stay (hospitalisation) Withdrawal severity (SIWA-Ar) Trainer experience Number of of relapses (any substance use during inpatient, AAT period) Treatment drop out (against medical advice during inpatient, AAT period) 6.2 Randomisation, blinding and treatment allocation After their first regular AAT session, participants will be assigned by coordinators per site on the basis of order in which they sign informed consent, to either the AAT-M or AAT-S condition by one of the appointed AAT coordinators using a predetermined randomization code order (using random blocksizes of 4, 6 and 8 to ensure the treatment groups are balanced at the end of every block), stratified per site and with an allocation ratio of 1:1 using Sealed EnvelopeTM (Sealed Envelope Ltd. 2021). Coordinators are staff members working at one of the three facilities and instruct patients and participants regarding the AAT. Principal investigators and research assistants administering the three and six month FU questionnaires will be blinded to allocation. After the last data is collected, the principal investigators receive the allocation code from the coordinators in order to analyse the data. 6.3 Study procedures Eligible participants will be identified in the pre-admission phase of inpatient treatment during weekly multidisciplinary staff meetings that discuss newly referred patients. If eligible, the AAT coordinator of the site contacts the patient after initial detoxification and provides information and instruction with regard to the AAT (as part of routine care) and the study. After informed consent, and after the baseline assessment and first, regular AAT session, participants will be assigned to the AAT-S or AAT-M condition following the randomization procedure outlined in 8.2. Participants in the AAT-S condition will receive one AAT session per week (a total of four sessions, 300 trails each). Participants in AAT-M condition will receive the same amount of AAT sessions and trials but massed within an 8 day period proximal to admission. After baseline assessment the AAT coordinator will plan the AAT sessions. Data collected from the assessments and AAT sessions will be uploaded to a Digital Research Environment by a research assistant. They also plan FU assessments (by telephone) with the participants 21 days + three and six months counted from the first AAT trail. 21 Days after the first trail both groups will have concluded all AAT trails. In practice, most patients are discharged after 4-6 weeks of inpatient treatment. During FU assessment, most if not all participants will have left inpatient treatment and will be receiving further outpatient treatment. Some participants may no longer receive addiction treatment at that time. However, they will still be approached for assessment. The assessments are carried out by AAT coordinators (baseline) and research assistants (FU). All AAT sessions are coordinated and guided in group format (a maximum of four patients/participants at once) by a regular staff member as it is part of TAU. If a session is cancelled, they are rescheduled by the staff member in the same week (if possible) or next week so that all sessions are delivered. After all FU data is collected by research assistants, the participant receives a voucher worth 15 euro, by mail. Instruments 1. Demographics and participant characteristics. At baseline, participants will be presented with some questions regarding demographics and personal characteristics: age, gender, highest attained educational level, marital status, ethnicity, social-economic status (source of income), type of job, relevant disabilities/diseases and medicine usage (see appendix 15.1). 2. Measurements in the Addictions for Triage and Evaluation (MATE) The MATE (version 2.1; Schippers & Broekman, 2011) is composed of 10 sections. The modules are made up of a set of existing short instruments from the public domain which are psychometrically solid as is evidenced in international published empirical studies. Conceptually, the MATE was constructed according to the ICD and International Classification of Functioning (ICF) in the World Health Organization (WHO) classification system. Two of the ICF-related modules were newly designed. In this study we only use parts 1, 3 and Q1 (see appendix 15.2). Section 1: 'USE'. 42 Questions to determine the severity of the use of psychoactive substances in the previous period as well as life-time. Inventarisation takes place in a grid whereby substances (selection derived from the CIDI vs 2.1 (World Health Organization, 1997), including nicotine and gambling), the level of use and typical form or route of application are reported. Use (in standard units) in the past 30 days and use on a typical day of use are reported, as well as the number of years of regular use. The primary problem substance or behaviour can be identified. Section 3: 'ADDICTION TREATMENT HISTORY'. Four Questions on how much previous addiction treatments the participant has undergone in the past 5 years. Section Q1: 'CRAVING'. These are five questions derived from the 14-items Obsessive-Compulsive Drinking (and drug use) Scale (OCDS: Anton, Moak, & Latham, 1996; OCDS5: de Wildt et al., 2005) which intend to measure core features of craving (timeframe: previous week). Together with the previous section, this section can be checked to see whether the participant has an AUD according to the DSM-5 criteria. 3. Measurements in the Addictions for Triage and Evaluation - Outcomes (MATE - Outcomes) The MATE - Outcomes (version 2.1; Schippers & Broekman, 2011) is the evaluation version of the MATE. However, the modules which are used in this study (1 and Q1) are exactly the same as those described above. 4. Duration of alcohol abstinence - Single additional question concerning the number of weeks of alcohol abstinence. Duration is described in terms of weeks until relapse starting from the point after AAT-S and AAT-M (plus 21 days) trails were completed. Number of weeks are expressed numerically where a value of '999' means the participant has remained abstinent. 5. Alcohol approach bias (training effect) The change in alcohol approach bias is not only an outcome, but also a predictor variable. It is computed using the 40 trials of the assessment phase at the beginning of each session. The bias score is derived by first computing, separately per patient and session, the mean reaction time (RT) of the four critical combinations of drink type and movement: pull-alcohol, push-alcohol, pull-soda, and push-soda. The score is then computed as a difference score by subtracting the two training-compatible combinations from the two training-incompatible combinations: (pull-alcohol + push-soda) - (push-alcohol + pull-soda). This way, more positive scores indicate more relative alcohol-avoidance. It is expected that on average, patients will initially show a negative score which will become more positive over the course of the training. 6.4 Withdrawal of individual subjects Participants can leave the study at any time for any reason if they wish to do so without any consequences. The treatment staff can decide to withdraw a participant from the study for urgent medical reasons or severe adverse (psychiatric) effects. The principal investigators can decide to withdraw a participant from the study in case of repeated failure to comply with AAT instructions, no-show, or showing up intoxicated at the start of a session. 6.5 Replacement of individual subjects after withdrawal Participants who have been randomized but are withdrawn (e.g. for medical reasons) or choose to withdraw before the start of the second AAT session are included according to intention to treat analysis. Enrolment will continue until 234 participants have been included. 6.6 Follow-up of subjects withdrawn from treatment Participants who withdraw prematurely from the treatment (study drop-outs) can contact the principal investigator, if questions arise later on. 6.7 Premature termination of the study Due to the non-invasive nature of the intervention this section is deemed not applicable. 7. SAFETY REPORTING 7.1 Temporary halt for reasons of subject safety Routine clinical care, including the AAT, is proven effective and applied in other countries, there are no safety concerns. Nevertheless, participants will be monitored and if they don't want to participate in AAT for whatever reason, they can stop at any time. 7.2 AEs, SAEs and SUSARs 7.2.1 Adverse events (AEs) Due to the non-invasive nature of the intervention this section is deemed not applicable. . 7.2.2 Serious adverse events (SAEs) Due to the non-invasive nature of the intervention this section is deemed not applicable. . 7.2.3 Suspected unexpected serious adverse reactions (SUSARs) Due to the non-invasive nature of the intervention this section is deemed not applicable. 7.3 Annual safety report Not applicable. 7.4 Follow-up of adverse events Not applicable. 7.5 [Data Safety Monitoring Board (DSMB) / Safety Committee] Not applicable. 8. STATISTICAL ANALYSIS Baseline differences in demographic and clinical characteristics are described in terms of numbers, percentages, means, standard deviations and ranges. A series of linear mixed models are performed using the R statistical program (R Core Team, 2019) with the lmer function in the lme4 package (Bates, Maechler, Bolker and Walker, 2015) to examine treatment group differences over time on all primary and secondary outcomes separately. In each model, time, treatment group, and the interaction between time and group are included as fixed effects; the intercept is treated as a random effect in order to account for individual differences in the levels of the outcome variable. An intention-to-treat (ITT) as well as a per-protocol analysis (PPA) will be carried out. In the ITT analysis all participants that have been allocated to one of the conditions (following randomization) will be included. In the PPA only those that have received all four AAT sessions are included. The primary outcome (the difference between the groups in changes in alcohol use (mean units of alcohol consumption in the past 30 days) between T0 and T2), is determined in both the ITT and the PPA. To determine p-values, Type 3 conditional F-tests are computed with the Kenward-Roger approximation for degrees of freedom. Effect sizes (partial eta-squared) and their corresponding 90% confidence intervals are obtained using the eta_sq function in the sjstats package (Lüdecke, 2019). Post hoc comparisons between groups per timepoint and between timepoint per group are done with Bonferroni adjustment. This will take the shape of planned contrasts whereby spaced > massed and T2-T0 > T2-T1 > T1-T0. 9. ETHICAL CONSIDERATIONS 9.1 Regulation statement The study has been designed and will be conducted according to the principles of the Declaration of Helsinki (version October 2008) and in accordance with the Dutch Medical Research Involving Human Participants Act (WMO). However, we believe the burden for participants is low since a significant part of the study aligns with TAU, in addition, the intervention is non-invasive, therefore we will seek a non-WMO declaration. 9.2 Recruitment and consent Potential participants that meet eligibility criteria are identified during multidisciplinary staff meetings. A staff member informs the eligible participants about the study, provides them with the patient information folder, and request them to read it. At the beginning of the detoxification (first week of admission), they are contacted by a staff member to see whether they are interested to participate, have any questions or need more information and are willing to give written informed consent. Participants can contact a staff member, the independent expert or a principal investigator when questions arise or when they want to withdraw their consent (contact details are provided within the written patient information form). 9.3 Objection by minors or incapacitated subjects (if applicable) Not applicable. 9.4 Benefits and risks assessment, group relatedness Participants receive routine clinical care. Additionally, one short questionnaire will be provided pre-AAT. Since all participant will be provided the same amount of AAT trails, albeit in different intervals, no adverse effects are to be expected. 9.5 Compensation for injury IrisZorg has a liability insurance which is in accordance with article 7 of the WMO. IrisZorg (also) has an insurance which is in accordance with the legal requirements in the Netherlands (Article 7 WMO). This insurance provides cover for damage to research subjects through injury or death caused by the study. The insurance applies to the damage that becomes apparent during the study or within 4 years after the end of the study. 9.6 Incentives (if applicable) Participants receive an incentive (a voucher worth € 15,-) after final FU assessment. 10. ADMINISTRATIVE ASPECTS, MONITORING AND PUBLICATION 10.1 Handling and storage of data and documents Data collected from the assessments and AAT sessions (reaction times) will be uploaded directly to a a secured cloud space, by a research assistant. They also plan FU assessments (by telephone) with the participants 21 days + three and six months counted from the first AAT trail. This data will also be stored at the secured cloud space. Analogue data (e.g. written informed consent forms) will be archived digitally in the cloud space. Routine care data such as DSM-5 diagnosis, length of stay (hospitalisation) and withdrawal severity (SIWA-Ar) data will be extracted from the electronic patient files after which it also will be entered into the secured cloud space. The data will be handled only by the researchers involved in the study and will be analyzed using SPSS. Each participant will be given an anonymous subject identification code based on order in which they signed their informed consent form, per site (Z = Zevenaar, W + Wolfheze and T= Tiel), e.g. Z001, Z002, Z003 etc.). Direct identifiable person-bound contact data (needed for follow-up data collection) will be kept (by the principal investigators) for the duration of the study. This information will be kept apart from the data. 10.2 Monitoring and Quality Assurance The research projects' studies conduct will not be monitored. 10.3 Amendments Amendments are changes made to the research after a favourable opinion by the accredited METC has been given. All amendments will be notified to the METC that gave a favourable opinion. 10.4 Annual progress report The investigator will submit a summary of the progress of the research project to the accredited METC once a year. Information will be provided on the date of inclusion of the first subject, numbers of subjects included and numbers of subjects that have completed the study, serious adverse events/ serious adverse reactions, other problems, and amendments. 10.5 Temporary halt and (prematurely) end of study report The investigator will notify the accredited METC of the end of the study within a period of 8 weeks. The end of the study is defined as the last patient's last follow-up assessment. In case the study is ended prematurely, the investigator will notify the accredited METC, including the reasons for the premature termination. Within one year after the end of the study, the investigator/sponsor will submit a final study report with the results of the whole study, including any publications/abstracts, to the accredited METC. 10.6 Public disclosure and publication policy All collected data will be submitted for publication in the form of a scientific article on the effectiveness of spaced versus massed AAT leaning. This article will be submitted to an international (English language) peer reviewed journal. In addition, results will be disseminated in the form of presentations of results and implications for clinical practice on symposia and congresses.

AAT-S participants will receive 4 AAT sessions totalling 1200 trails but spaced out over four weeks (one AAT session per week)

AAT-M participants will receive 4 AAT sessions (each with 300 trials) within the space of 8 days (which totals 1200 trails).

Participants are instructed to react with a joystick to the tilt of pictures, presented on a computer screen. For instance, participants make an approach movement (pull) with left-tilted pictures, and make an avoidance movement (push) with right-tilted pictures. After the correct movement is made, the pictures disappear, otherwise they do not, until the correct response is made.

o Changes in mean units of alcohol consumed in the past 30 days (MATE/MATE - Outcomes (Schippers, Broekman & Buchholz, 2007) section 1: number of regular drinking days x average number of drinks regular drinking day + number of heavy drinking days x average number of drinks heavy drinking day)

o Changes in mean craving score over the past 7 days (MATE/MATE - Outcomes section 'Q1' from the OCDS-5 (Schippers et. al, 1997): thoughts and desires concerning wanting to use alcohol

Inclusion Criteria: A primary diagnosis of alcohol use disorder (meeting the DSM-5 criteria (American Psychiatric Association, 2013); Age of at least 18 years or older; Good Dutch language proficiency; Written informed consent. Exclusion Criteria: Severe, current psychiatric symptoms (especially manic, psychotic, suicidal and aggressive symptoms) that may endanger participants or others and jeopardize study adherence (determined on a case-by-case basis. The staff member who will provide the AAT information, informed consent and instruction will always consult one of the principal investigators in that case).

● Aharonovich, E., Shmulewitz, D., Wall, M. M., Grant, B. F., & Hasin, D. S. (2017). Self-reported cognitive scales in a US national survey: reliability, validity, and preliminary evidence for associations with alcohol and drug use. Addiction, 112(12), 2132-2143. ● Brandon, T. H., Vidrine, J. I., & Litvin, E. B. (2007). Relapse and relapse prevention. Annu. Rev. Clin. Psychol., 3, 257-284. ● Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological bulletin, 132(3), 354. ● Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning. Psychological science, 19(11), 1095-1102. ● Domínguez-Salas, S., Díaz-Batanero, C., Lozano-Rojas, O. M., & Verdejo-García, A. (2016). Impact of general cognition and executive function deficits on addiction treatment outcomes. Neuroscience & Biobehavioral Reviews, 71, 772-801. ● Donovan, J. J., & Radosevich, D. J. (1999). A meta-analytic review of the distribution of practice effect: Now you see it, now you don't. Journal of Applied Psychology, 84(5), 795. ● Eberl, C., Wiers, R. W., Pawelczack, S., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2013). Approach bias modification in alcohol dependence: do clinical effects replicate and for whom does it work best?. Developmental Cognitive Neuroscience, 4, 38-51. ● Eberl, C., Wiers, R. W., Pawelczack, S., Rinck, M., Becker, E. S. & Lindenmeyer, J. (2014). Implementation of approach bias re-training in alcoholism - how many sessions are needed? Alcoholism: Clinical and Experimental Research, 38, 2, 587-594. ● Greene, R. L. (1989). Spacing effects in memory: Evidence for a two-process account. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15(3), 371. ● Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General, 110(3), 306. ● Kakoschke, N., Kemps, E., & Tiggemann, M. (2017). Approach bias modification training and consumption. Addictive behaviors, 64, 21-28. ● Manning, V., Staiger, P. K., Hall, K., Garfield, J. B., Flaks, G., Leung, D., ... & Verdejo-Garcia, A. (2016). Cognitive bias modification training during inpatient alcohol detoxification reduces early relapse: a randomized controlled trial. Alcoholism: Clinical and Experimental Research, 40(9), 2011-2019. ● Satpute, A. B. & Lieberman, M. D. (2006). Integrating automatic and controlled processes into neurocognitive models of social cognition. Brain Research, 1079 (1), 86-97. ● Schneider , W. & Chein, J. M. (2003). Controlled & automatic processing: behavior, theory, and biological mechanisms. Cognitive Science, 27, 525-559. ● Sherman, J. W., Gawronski, B., Gonsalkorale, K., Hugenberg, K., Allen, T. J. & Groom, C. J. (2008). The Self-Regulation of Automatic Associations and Behavioral Impulses. Psychological Review, 115(2), 314-335. ● Smith, E. R. & DeCoster, J. (2000). Dual-Process Models in Social and Cognitive Psychology: Conceptual Integration and Links to Underlying Memory Systems. Personality and Social Psychology Review, 4(2), 108-131. ● Stacy, A.W. & Wiers R.W. (2010). Implicit Cognition and Addiction: A Tool for Explaining Paradoxical Behavior. Annu Rev Clin Psychol (6), 551-575. ● Strack, F. & Deutsch, R. (2004). Reflective and Impulsive Determinants of Social Behavior. Personality and Social Psychology Review, 8(3), 220-247. ● Wiers, R.W., Bartholow, B.D., van den Wildenberg, E, Thush, C., Engels, R.C.M.E., Sher, K.J., Grenard, J., Ames, S.L., Stacy, A.W. (2007). Automatic and controlled processes and the development of addictive behaviors in adolescents: a review and a model. Pharmacological Biochemical Behaviour, 86, 263-283. ● Wiers, R.W., Van Woerden, N., Smulders, F.T.Y., De Jong, P.J. (2002): Implicit and explicit alcohol-related cognitions in heavy and light drinkers. Journal of Abnormal Psychology, 111, 648-658. ● Wiers R.W., Eberl, C., Rinck, M., Becker, E.S., & Lindenmeyer, J. (2011). Retraining automatic action tendencies changes alcoholic patients' approach bias for alcohol and improves treatment outcome. [Internet]. Psychological Sciences, 22, 490-497. ● Wiers R.W., Gladwin, T.E., Hofmann, W., Salemink, E., Ridderinkhof, K.R. (2013). Cognitive Bias Modification and Cognitive Control Training in Addiction and Related Psychopathology: Mechanisms, Clinical Perspectives, and Ways Forward. Clinical Psychological Sciences, 1, 192-212. ● Wiers, R. W., Rinck, M., Kordts, R., Houben, K., & Strack, F. (2010). Retraining automatic action-tendencies to approach alcohol in hazardous drinkers. Addiction, 105, 279-287. ● Wiers, R.W., Rinck, M., Dictus, M., & van den Wildenberg, E. (2009). Relatively strong automatic appetitive action-tendencies in male carriers of the OPRM1 G-allele. Genes Brain Behav., 8(1), 101-6.