Background: This systematic review and meta-analysis aims at assessing the composition and performance of care management models evaluated in the last decade and their impact on patient important outcomes. Methods. A comprehensive literature search of electronic bibliographic databases was performed to identify care management trials in type 2 diabetes. Random effects meta-analysis was used when feasible to pool outcome measures. Results: Fifty-two studies were eligible. Most commonly reported were surrogate outcomes (such as HbA1c and LDL), followed by process measures (clinic visit or testing frequency). Less frequently reported were quality of life, patient satisfaction, self-care, and healthcare utilization. Most care management modalities were carved out from primary care. Meta-analysis demonstrated a statistically significant but trivial reduction of HbA1c (weighted difference in means -0.21%, 95% confidence interval -0.40 to -0.03, p <.03) and LDL-cholesterol (weighted difference in means -3.38 mg/dL, 95% confidence interval -6.27 to -0.49, p <.02). Conclusions: Most care management programs for patients with type 2 diabetes are 'carved-out', accomplish limited effects on metabolic outcomes, and have unknown effects on patient important outcomes. Comparative effectiveness research of different models of care management is needed to inform the design of medical homes for patients with chronic conditions. © 2012 Egginton et al; licensee BioMed Central Ltd.