Supplement 2: Treatment fidelity
To assure intervention fidelity, several actions were taken [53]. To promote intervention fidelity session-to-session therapist manual and adolescent workbook were written for both treatments, to work with during sessions. Therapists were trained in both treatments in a two-day training, which was given by the treatment developers and member of Motivational Interviewing Network of Trainers. Since this within-therapist design enhances the chance of spillover effects, during this training similarities and differences between treatments were emphasized. Also, role-playing was used for practice and videos were showed to model both treatments. Every other week therapists had supervision to discuss ongoing treatments. Eight participating supervisors, acknowledged by the Dutch Association for Behavioral and Cognitive Therapies, were trained by the treatment developers and a MINT member (for 3 days), specifically on supervision of MI and both treatments. During supervision role-play was used to practice problem situations and films of therapists’ treatments were watched to monitor treatments and provide corrective feedback whenever needed. A year after training all therapists attended a one-day booster training where again similarities and differences between treatments were emphasized and in addition, problems/pitfalls that came forward in supervision were discussed.
Measures of intervention fidelity
Filmed adherence and contamination Sessions of both treatments (8 adolescent- and 2 parental sessions) were filmed. To rate adherence and contamination, films were randomly chosen stratified on session and on treatment condition. Five films were rated of every session of every treatment condition, resulting in 100 rated sessions (i.e. 5.4% of all sessions).
An alphabetical list of all elements of both treatments was created for rating. Four master level psychology students were trained by the treatment developers to recognize all treatment elements of both treatments, while being blind to which treatment the elements belonged. Raters were also blind to the session number and treatment of the films they rated. The four students had a high inter rater reliability of ICC(1, k)=.97. Adherence is measured by calculating the percentage of elements that are addressed in the right session of the right treatment. Contamination is measured by calculating the percentage of elements that are addressed in one treatment while actually belonging in the other treatment. Adherence and contamination were calculated for adolescent and parent sessions separately.