199 Summary and General discussion 7 informal social network intervention in forensic psychiatric patients, is considered the main strength of our research. Another important strength of this dissertation is the combination of rigorous and mixed methods that were used to examine effectiveness. We conducted a comprehensive meta-analysis using a multilevel approach, which enabled us to include all outcomes of studies and achieve maximum statistical power (Assink & Wibbelink, 2016). Furthermore, we conducted an RCT with a relatively large forensic population who participated in multiple measurements over a period of 18 months. Data collection with multiple measurement over time resulted in higher statistical power. Additionally, the qualitative part of this research helped to access perspectives of patients and volunteers, which provided meaning to the quantitative results of the RCT. Lastly, the findings in this dissertation are encouraging for clinical practice, as the research was implemented in clinical practice for forensic psychiatric outpatients, enhancing the ecological validity. Moreover, the commitment of multiple stakeholders (i.e., scientist-practitioners, informal care staff, clinicians, and management) contributed to a sustainable implementation of the informal social network intervention in clinical practice. This dissertation also has several limitations worth noting. First, given the rigorous methods that were used in chapter 2 and 5 (i.e., the meta-analysis and RCT, respectively), results might not be generalizable to individual patients (Purgato et al., 2021). The meta-analysis is a broad study, including different types of social network interventions and psychiatric populations. Despite our attempt to determine the effects of specific intervention characteristics as well as subgroups of patients, our findings did not yet disentangle which type of intervention should be recommended and which subgroup of patients could benefit. Furthermore, the RCT was conducted among forensic psychiatric outpatients, consisting of a wide range of patients with various psychiatric disorders. Although moderator analyses did reveal different effects across subgroups of patients (i.e., males versus females, patients with primary substance use disorders and comorbidity), these effects should be considered exploratory and preliminary. Additionally, in chapter 4 and 5 we encountered high levels of intervention dropout as well as different levels of compliance, both of which likely affected treatment effectiveness. Therefore, especially since this is the first study, we cannot draw firm conclusions about treatment effects in forensic psychiatric outpatients. However, we do know that treatment effects of the per-protocol analyses, in which no, low, and high compliance groups were
RkJQdWJsaXNoZXIy MjY0ODMw