Exploratory cluster randomised controlled trial of shared care development for long-term mental illness.

BACKGROUND: Primary care clinicians have a considerable amount of contact with patients suffering from long-term mental illness. The United Kingdom's National Health Service now requires general practices to contribute more systematically to care for this group of patients. AIMS: To determine the effects of Mental Health Link, a facilitation-based quality improvement programme designed to improve communication between the teams and systems of care within general practice. Design of study: Exploratory cluster randomised controlled trial. SETTING: Twenty-three urban general practices and associated community mental health teams. METHOD: Practices were randomised to service development as usual or to the Mental Health Link programme. Questionnaires and an audit of notes assessed 335 patients' satisfaction, unmet need, mental health status, processes of mental and physical care, and general practitioners' satisfaction with services and beliefs about service development. Service use and intervention costs were also measured. RESULTS: There were no significant differences in patients' perception of their unmet need, satisfaction or general health. Intervention patients had fewer psychiatric relapses than control patients (mean = 0.39 versus 0.71, respectively, P = 0.02) but there were no differences in documented processes of care. Intervention practitioners were more satisfied and services improved significantly for intervention practices. There was an additional mean direct cost of pound 63 per patient with long-term mental illness for the intervention compared with the control. CONCLUSION: Significant differences were seen in relapse rates and practitioner satisfaction. Improvements in service development did not translate into documented improvements in care. This could be explained by the intervention working via the improvements in informal shared care developed through better link working. This type of facilitated intervention tailored to context has the potential to improve care and interface working.