Participants were recruited within a randomized controlled trial (RCT) in Stockholm, Sweden, that investigated the efficacy of two psychological treatments: Behavioral Self-control Training (BSCT) and MET, including 250 individuals with AUD and a goal of controlled drinking. The primary outcome of the trial has been reported elsewhere. For detailed information on participants and study procedures in the RCT, see Ingesson Hammarberg, in press. The original trial was approved by the Regional Ethics Board in Stockholm (DNR: 2016/634-31/2). The current study was approved by the Regional Ethics Board, as an amendment to the original study. The trial was retrospectively registered in ISRCTN (14,539,251) (05/06/2018). The study was conducted in accordance with the Declaration of Helsinki and was reported in accordance with the Consolidated criteria for Reporting Qualitative research) (COREQ) checklist [21].

All patients underwent assessment as part of the clinical trial protocol, including of level of alcohol consumption, alcohol-related consequences, and psychiatric diagnostics. Inclusion criteria were: a stated goal of controlled drinking, fulfilment of a diagnosis of alcohol use disorder, and age 18–70 years. Exclusion criteria were: fulfilment of any other substance use disorder except nicotine, severe psychiatric comorbidity, and severe somatic risk related to continuous alcohol consumption.

Participants in this study were recruited during 2021, among those who had received MET and attended the 26-week follow-up in the RCT. If consenting, participants took part in an interview either via a video meeting platform or by telephone. All participants gave verbal and written consent to participate and to data being used in a scientific publication. The interviews were audio-recorded and then transcribed by the interviewers. None of the participants withdrew consent after being interviewed and there was no reimbursement for their participation.

MET, as applied in this study, included four sessions, and was distributed over 12 weeks [15]. Patient assessment was conducted by an assessor as part of trial routines. Thereafter, at the first treatment session, the assigned therapist offered feedback on this assessment. Due to the COVID-19 pandemic, the majority of participants received their treatment sessions via video meetings. The manual also included two worksheets with the themes (1) change plan and (2) how to maintain change. These could be used within a session or as homework assignments.

Five MET therapists were involved in the current study. All had extensive training in MI and MET, with clinical experience in both MI and MET ranging from 10 to 20 years. Therapists attended regular supervision meetings and received regular feedback on their recordings in accordance with the Motivational Interviewing Treatment Integrity Code (MITI) protocol 4.2.1 [22]. All sessions of which therapists had patients’ consent were recorded to assess treatment integrity in the conducted treatments. A total of 10% of the recorded MET sessions were randomly chosen and scored in line with the Motivational Interviewing Treatment Integrity code (MITI) 4.2.1 by an external expert team. The four areas which were evaluated were; (1) Technical score (average of cultivating change talk + softening sustain talk/2); (2) Relational score (average of empathy + partnership/2); (3) Percentage of complex reflections of all reflections and; (4) Reflection to question ratio) [22].

Baseline measures, as well as sociodemographic data, were collected as part of the clinical trial protocol. Alcohol consumption was assessed using the timeline follow-back method [23] over a period of 90 days, in number of standard drinks per day (12 g of pure ethanol) at baseline and follow-up.

A semi-structured interview guide was developed by the research team. Four areas were covered: (1) The treatment experience, and if there were positive and negative features. (2) Whether the treatment was helpful and, if so, in what way? If not, in what way? (3) Was the treatment sufficient to achieve change regarding alcohol consumption? To what extent? In what way? (4) Thoughts about involving a significant other (SO) in the treatment. The guide was discussed after the first interview and was not changed thereafter.

The interview guide allowed for participants to freely express opinions and permitted follow-up questions from the researcher. The approach to data was inductive, as research questions and interview guide were not chosen based on a specific theory or concept that we wanted to explore. Rather, they were based on our interests and clinical experiences on how patients perceived the therapeutic intervention.

Interviews were carried out and transcribed by two of the authors (JS and RT). Both are female, trained clinicians (nurse and psychologist), well-acquainted with the practices of MET, but neither was a therapist at the clinic where treatments were conducted. Authors JS and SIH (both female) were both active as research coordinators in the RCT, and acquainted with the individuals who participated in the study. Authors SIH (female) and AH (male) are both trained MI therapists, but were not active as therapists in the current project. The authors’ pre-understanding of MI/MET as clinicians at the treatment site could potentially have had an impact on the analysis. The two-legged position as both therapists and researchers was key in critically evaluating and interpreting data. Lastly, author CN (female) is a senior researcher with long experience of qualitative research. CN is affiliated with a different university and was therefore not involved in the data collection process, or active as a therapist at the clinic, which allowed for a more distanced view of the data.

Thematic analysis in accordance with the six phases defined by Braun and Clarke [24, 25] was used in the current study. Transcripts were read independently and repeatedly by the authors, with the research questions in mind. Sentences and meaning units in line with the aims of the study were coded and, after joint discussion, sorted into preliminary themes. The material was re-read and themes were reviewed. Analysis continued until all themes were deemed to be clearly defined and distinct from one another. All authors discussed the coding of data until a consensus was reached and themes were perceived as concisely describing the content. Once 15 interviews had been conducted, information from the final ones did not produce any changes to the themes. Therefore, saturation was deemed to have been achieved [26].