Group Version of CRAFT Compared to Self-Directed CRAFT Delivery and Non-intervention: a Three-armed RCT

The aim of this study is to determine the applicability and effectiveness of (a) CRAFT group format in conjunction with the CRAFT self-help book and (b) a CRAFT self-help book (self-directed CRAFT delivery condition), compared to (c) non-intervention (control). The primary aim is increase CSO well- being and reduce CSO health complaints. CSOs will be assessed at baseline (T0), 2 months (end of intervention) (T1), 3 months- (T2), and 6 months follow-up (T3), independently of the assigned condition. Secondary outcomes will encompass cost-effectiveness of both CRAFT interventions. Also, we measure IP engagement and IPs' substance use.

Since the negative consequences of substance abuse are not limited to alcohol- and drug users themselves, it also affects the lives of family members and close friends (CSOs). This study focusses on CSOs to improve their lives. Recruitment will take place in the Netherlands (nation- wide). Assessment and the CRAFT group intervention will be conducted at six different locations (Arnhem, Doetinchem, Ede, Nijmegen, Tiel and Zevenaar, all in the province of Gelderland) in the Netherlands. Participants will be recruited via advertisements on the Internet, social media, websites of IrisZorg (centre for addiction care and sheltered housing ), and subsequently, in local newspapers and flyers located in emergency rooms, general practitioners, family practice clinics, and mental health based treatment agencies.

To obtain information about the presence of psychiatric disorders of the CSOs, a Dutch version of the Mini-International Neuropsychiatric Interview (M.I.N.I.) will be used. After initial assessment, a randomization procedure will be conducted by an independent researcher, who is affiliated to the Erasmus Medical Center, Rotterdam, the Netherlands. CSOs will be informed about the allocation of conditions within 1 week, after initial admission. Several outcome measures will be used in this study. The following will be measured:

1. CSO well- being (including physical and psychological health and complaints, quality of life, quality of the relationship, self- efficacy and social support);
2. cost-effectiveness of both CRAFT and Self- Directed Delivery compared to non- intervention.
3. IP engagement and IP substance use.

Using the primary outcome, CSO well-being, a priori power analysis is completed. Means and standard deviations for CSO well-being, measured with the Beck Depression Inventory (BDI), for Group CRAFT were used (Miller et al., 1999). An effect size was calculated (f = 0.22). The power analysis, conducted with the online program called G-power, indicated that a total sample size of 69 CSOs is needed to achieve a power of 95% for detecting a medium effect size when using the p = 0.05 criterion of statistical significance. These numbers are based on an ANOVA with a within-between subject design. Based on the pilot study of Kirby et al. (1999), 23% drop out after 3- and 6- month follow-up is taken into account. This means that, including the 23% drop-out-rate and based on the sample size of 69, an N of 85 is needed after 3 months and that a total N of 105 is needed to complete all assessments. Summarized, to be able to have at least 69 assessments completed at 6- month follow- up, a total sample size of 105 CSOs is required; with an estimated total drop-out rate of 34%.