Abstract
Background/Aims Individuals with serious mental illnesses can experience significant disability, affecting ability to guide and adhere to treatment and navigate systems. Needs for a range of services are common, yet despite recent efforts to improve care quality, persistent challenges blunt or derail reforms. Obstacles include: difficulties changing the focus of care from acute symptom control/relapse prevention to patient-centered/recovery-focused care; constructing coherent, integrated services; coordinating care while managing multiple, fluctuating funding streams; and high staff turnover that produces poor continuity of care. Our goal is to provide guidance for future comparative effectiveness and patient-centered outcomes research (CER, PCOR) to improve individual-level patient-centered outcomes (PCOs).
Methods With stakeholders and technical experts, using an iterative consensus approach, we created a definition of what constitutes a care and service delivery intervention, constructed a theoretical model of a learning system to improve quality and coordination of care, developed an analytic framework, and conducted a narrative review to identify research gaps, high leverage points, and key questions for future CER and PCOR.
Results (1) We need more patient centered outcomes developed by or in concert with service users. (2) Information regarding the outcomes service users value most and least remains limited. Value-based information is fundamental to patient-centered care and PCOs. (3) Most efforts to improve quality of mental health services have focused on care processes and necessary institutional structures. Links between indicators of process and structure, and PCOs, are nearly non-existent. (4) Few efforts have been made to aggregate PCOs to provide performance feedback at the clinician, organization, or system level; methods and processes are needed. (5) Financing of services is structured in ways that complicate and impede coherent, integrated delivery, and research comparisons. (6) Current CER information is not adequate to produce system change. Complexity theory suggests focusing on organizational culture/climate, supporting employees and promoting high-quality interactions/teamwork. Interactions are the nexus of information processing/sense-making that are necessary in a learning system.
Conclusions Health care reform is creating multiple opportunities to exceed current incremental efforts to improve outcomes. Developing learning systems that provide real patient-centered/patient-directed care to individuals with serious mental illnesses should capitalize on these openings.

