The plan is to conduct the study through a cluster-randomized approach. Recruitment will take place amongst inpatients in the Swedish forensic mental health services. The clusters of patients randomized to the two study conditions will in effect be inpatient forensic mental health wards at Swedish forensic facilities. Enrolment and randomizing of clusters will take place stepwise during the duration of the study. Inpatient wards at participating sites will be randomized to either the control or the active condition and patients at these wards will be asked to give informed consent before participating in the study. After training, the staff at these wards will administer IMR to patients during two group sessions and one individual session a week. The IMR treatment administered consist of 10 chapters containing information and exercises pertaining to subjects relevant to mental health and recovery, such as medication use, social support, problem solving, etc. Translation of the original IMR-manual into Swedish and adaptions to the forensic mental health setting was done with permission from the original developers and authors of the treatment protocol. Study participants at control wards will receive treatment as usual. Data collection takes place at four times during the study (before start of treatment, four months into treatment, post treatment and during a three month follow-up). During data collection, participants fill in self-report measurements of symptom levels, functional impairment and perceived hope. Clinicians with good knowledge of participants also complete informant questionnaires on perceived levels of patient symptoms and functioning. During the data collection that takes place before treatment, participants also complete a neuropsychological test ("Zoo park" from the Behavioral assessment of dysexecutive syndrome, BADS). Anamnestic data pertaining to the pre-admission functioning and criminal history of the patients in the active condition will also be collected from the patient files. The data collection mentioned in the previous two sentences will be done to investigate potential moderators of treatment outcomes.
The outcome measures used in the study are WHODAS 2.0 (World Health Organization Disability Schedule 2.0 self-report and informant versions), IMRs (Illness management and Recovery treatment scale, self-report and informant versions), HoNOS-S (Health of Nation Outcome Scale-Secure version) and ASHS (Adult State Hope Scale). During the data collection four months into treatment, only IMRs and ASHS is completed, whereas all outcome scales are completed during the other data collection points. During the data collection post-treatment, treatment participants also complete CSQ-8 (Client Satisfaction Questionnaire-8).
The experiences of the staff administrating the treatment also will be investigated, with the purpose of exploring the process of implementing the intervention in this setting. This will be done using the Swedish version of the NoMAD-questionnaire (Normalization Process Theory Measure), which is a tool developed to operationalize the theoretical framework of Normalization Process Theory. The staff will complete this questionnaire after receiving training in the intervention, three months after treatment commencement and during a follow-up a year after treatment commencement. During the latter two of these data collection points staff will also be interviewed in a semi-structured fashion, using the questions in the NoMAD-questionnaire as a starting point for collection of more in depth qualitative data on the implementation of IMR. The part of the study pertaining to staff experiences of working with IMR are exploratory in character and we do not consider NoMAD to be a measure of treatment outcome.
As mentioned above, certain anamnestic data and neuropsychological performance (measured by the Zoo-map subtest), will be collected and analyzed as potential moderators of treatment outcome. The pre-treatment self and informant-report of everyday functioning (measured by WHODAS 2.0) will also be analyzed as a potential moderator of treatment effects on the other outcome measures. Of special interest in this regard will be discrepancies between the participants´ self reported functional status (assessed by WHODAS 2.0 self report) and other measures of functional status (informant version of WHODAS 2.0, performance on the Zoo map-test and anamnestic indicators of functioning). This interest is motivated by previous research indicating that overestimation of own abilities play a part in predicting outcomes for schizophrenia spectrum patients.
