BackgroundWhile integrated primary healthcare for the management of depression has been well researched, appropriate models of primary care for people with severe and persistent psychotic disorders are poorly understood. In 2010 the NSW (Australia) Health Department commissioned a review of the evidence on "shared care" models of ambulatory mental health services. This focussed on critical factors in the implementation of these models in clinical practice, with a view to providing policy direction. The review excluded evidence about dementia, substance use and personality disorders.MethodsA rapid review involving a search for systematic reviews on The Cochrane Database of Systematic Reviews and Database of s of Reviews of Effects (DARE). This was followed by a search for papers published since these systematic reviews on Medline and supplemented by limited iterative searching from reference lists.ResultsShared care trials report improved mental and physical health outcomes in some clinical settings with improved social function, self management skills, service acceptability and reduced hospitalisation. Other benefits include improved access to specialist care, better engagement with and acceptability of mental health services. Limited economic evaluation shows significant set up costs, reduced patient costs and service savings often realised by other providers. Nevertheless these findings are not evident across all clinical groups. Gains require substantial cross-organisational commitment, carefully designed and consistently delivered interventions, with attention to staff selection, training and supervision. Effective models incorporated linkages across various service levels, clinical monitoring within agreed treatment protocols, improved continuity and comprehensiveness of services.Conclusions"Shared Care" models of mental health service delivery require attention to multiple levels (from organisational to individual clinicians), and complex service re-design. Re-evaluation of the roles of specialist mental health staff is a critical requirement. As expected, no one model of "shared" care fits diverse clinical groups. On the basis of the available evidence, we recommended a local trial that examined the process of implementation of core principles of shared care within primary care and specialist mental health clinical services.