General discussion 191 functional and personal recovery were stronger than the indirect associations via clinical recovery. Moreover, there were no associations with other components of illness management, including insight, medication adherence, addiction, and any type of recovery (Chapter 4). As mentioned above, IMR was found to affect overall self-reported illness selfmanagement and personal recovery. Mediation analysis using longitudinal data showed that the direct effects of overall illness self-management on personal recovery were stronger than the indirect effects through improvement in clinical and functional recovery (Chapter 7). This suggests that the found effects of IMR may be directly interrelated. However, a mediation analysis did not prove causality. The results of the mediation analysis only partly supported the working of IMR, as suggested in the conceptual framework originally described by Mueser (1). Moreover, based on the results described in Chapter 7, a possible new variant of the IMR framework was presented, suggesting that functional recovery is a mediator between illness management and personal recovery. Both analyses (Chapter 4 and 7) suggested that—different from the original conceptual framework—clinical recovery might not be a prerequisite for functional or personal recovery. However, both analyses, consistent with a meta-analysis (2), support the relevance of self-management interventions such as IMR for personal recovery. Comparing results with other studies Comparing effects found in our study with those of prior RCTs Six earlier RCTs on IMR yielded inconsistent results (3). Although the positive effect on our primary outcome—illness self-management measured using the client version of the IMR scale—was consistent with three previous RCTs on IMR (4-6), it differed from those of three other RCTs (7-10). Moreover, our negative results regarding the clinician version of the IMR scale are consistent with those of one previous RCT (9, 10); however, they differ from the results of four previous RCTs that reported positive outcomes (4-7, 11). Our positive findings regarding personal recovery differ from those of three previous RCTs (5, 8-10). Three earlier RCTs reported that IMR had positive effects on reducing psychiatric symptoms (4-6), whereas our negative results on symptoms were consistent with three RCTs (4, 8, 10, 12). Regarding hospitalization, only one RCT reported positive results (6); similar to our RCT, the remaining RCTs reported null results (4, 5, 7, 8, 12) (Table 2).
RkJQdWJsaXNoZXIy MjY0ODMw