Determinants of personal recovery: a mediation analysis 181 Our RCT on IMR , from which the data of this mediation analysis are derived, was conducted in a natural setting (11). Therefore, we suggest that the results of the current analysis may be well generalizable. Limitations This mediation analysis was based on difference scores of all four concepts on two time points; this study, therefore, used longitudinal information, but we were not able to infer causality. This would require more measurement points of time than those in this study. Future research should include longitudinal analysis. Another limitation is that all outcomes used were self-report questionnaires, which are inherently subjective. This may have caused bias in the associations between the measured concepts. Therefore, future research can also use clinician-rated outcomes, such as the clinician-rated IMR-Scale measuring illness self-management (30), the Brief Psychiatric Rating Scale for clinical recovery (44), and the Social and Occupational Functioning Assessment Scale for functional recovery (45). Personal recovery should be self-reported. In the present study, the MHRM was used, which is a composite personal recovery scale. However, as there are various measures for personal recovery (22), future research could use another scale for measuring personal recovery, such as the Recovery Assessment Scale (46) Conclusions This study’s results showed empirical support for the association between illness self-management and personal recovery. Moreover, self-reported illness management appeared to be a stronger direct determinant of personal recovery than indirectly via clinical and functional recovery. Furthermore, this study confirms that improvement in clinical and functional recovery might not be a prerequisite for improvement in personal recovery. Therefore, this study supports the relevance of self-management interventions such as IMR. These mental health services deserve more attention and should be part of the standard mental health care for people with SMI. The results partly provide empirical support for the working of IMR, as suggested in the conceptual framework. However, in this study, a possible new variant of this IMR framework was demonstrated, because functional recovery could be a mediator between illness management and personal recovery. To further investigate our results, future research using a longitudinal analysis and other outcome-measures for measuring the same concepts is recommended.
RkJQdWJsaXNoZXIy MjY0ODMw