This study will develop and evaluate a complex intervention to implement guidelines on family involvement for patients with psychotic disorders (F20-29 in International Classification of Diseases ICD-10) in community mental health centres, by using a cluster randomised design. Fifteen Norwegian outpatient units participate in the study, and each of them constitutes a single cluster, except for two collaborating clinics who are considered one cluster.

Of the fourteen clusters, half will receive implementation support and training immediately, whereas the other half will receive it one and a half year later. The study will assess both service level outcomes, by measuring fidelity scores, and selected outcomes for patients and relatives, by collecting questionnaires and data from central health registers and patient records. In addition, qualitative interviews will be performed with patients, relatives and health care personnel. The study will also include a cost-effectiveness analysis and a political economy analysis.

Background:

Family involvement during severe mental illness, such as psychotic disorders, is both important and challenging. Evidence suggest that family interventions for persons with psychotic disorders are associated with positive outcomes for both relatives and patients, and economic analyses of such interventions consistently report net saving. There are also important moral imperatives to involve those providing unpaid and informal care. Yet research has shown that relatives of patients with severe mental illness experience little involvement, and that the implementation of family interventions is patchy. The Norwegian national guidelines on family involvement in the public health- and care services and the national guidelines on psychotic disorders, both give recommendations on family involvement. However, there is little knowledge about how to achieve their implementation, and whether a high degree of implementation will be associated with improvements in selected outcomes for patients and relatives.

Setting:

Fifteen outpatient units from community mental health centres in the South-Eastern Norway Regional Health Authority.

Research questions:

1. What is the current level of implementation of the selected recommendations in the national guidelines on family involvement for persons with psychotic disorders in participating clinical units?
2. What are important barriers to and facilitators for implementing the national guidelines among the stakeholders at the clinical, organisational, and policy level?
3. What are important moral dilemmas and conflicting interests related to family involvement, and how can these be resolved?
4. Is implementation of the selected recommendations increased by a comprehensive implementation support program, compared with no such support?
5. Is a higher level of implementation of the selected recommendations associated with improvements in selected outcomes for patients and relatives?
6. Is implementation of family involvement during primary psychotic disorders a cost-effective intervention?

Hypotheses:

1. The current implementation of the selected recommendations in the national guidelines on family involvement for persons with psychotic disorders is low.
2. There are important barriers to and facilitators for implementing the national guidelines among the stakeholders, at the clinical, organisational and policy level.
3. There are important moral dilemmas and conflicting interests, and these can be dealt with through systematic triadic approaches and ethics reflection.
4. A comprehensive implementation program for the selected recommendations is associated with a significantly higher implementation of family involvement for persons with psychotic disorders, compared to no such specific program.
5. Higher implementation of the selected recommendations is associated with improved outcomes for patients and relatives.
6. Outcomes for relatives, patients and the public health- and welfare services justify the costs of implementing family involvement for persons with psychotic disorders.

Eligibility criteria apply to participants in both quantitative and qualitative sub-studies, except that participation in family psychoeducation as described below is NOT an exclusion criterion in the qualitative sub-study.

Inclusion criteria for patients:

• To have an established psychotic disorder (F20-29 in ICD-10) or a tentative diagnosis of psychotic disorder, certain enough to begin treatment. This need not be the patient's primary diagnosis.
• To be 18 years or older at the time of inclusion.

Exclusion criteria for patients:

• To be sentenced to psychiatric treatment.
• Not being competent to consent to participation in research.
• Having completed more than five joint sessions of family psychoeducation in single-family groups (patient and relative together) or more than ten joint sessions (multiple families together) in multiple-family groups, or a similarly structured family intervention.
• Not having any relatives or next of kin.

Inclusion criteria for relatives:

• Being a relative of a patient with a diagnosis as described above.
• To be 18 years or older at the time of inclusion.

Exclusion criteria for relatives:

• Having completed more than five joint sessions (patient and relative together) of family psychoeducation in single-family groups or more than ten joint sessions (multiple families together) in multiple-family groups, or a similarly structured family intervention.

Patients and relatives will not be informed about their outpatient clinic's allocation status. However, they may deduce this from the kind of treatment they receive. It's impossible to blind the care providers and for practical reasons we won't be able to blind the investigators or the outcomes assessors.