OBJECTIVES: Integrated care is increasingly being adopted as an effective approach to reduce fragmentation and promote comprehensive delivery of health services, and shows promise in improving outcomes for people with multimorbidity (MM), for which the burden is high globally. Decision Analytic Models (DAM) provide a systematic approach to evaluate the long-term health economic impacts of health interventions beyond those obtained within trial contexts. This systematic review examined the application of DAMs for the economic evaluation of integrated care models for patients with cardiometabolic MM.

METHODS: We conducted a systematic search in Medline, Web of Science, EMBASE, CINAHL, APA Psychinfo, Econlit, Scopus, and the Cochrane databases, published in English language until December 2023. We summarised the types of DAMs developed, the type of integrated care models evaluated, the MM diseases of focus, the impact of the interventions, and appraised the quality of reporting using a standard checklist.

RESULTS: We identified 16 full economic evaluations between 2006 and 2021. Majority (n=13) performed cost utility analysis (81%) and were conducted in high-income countries (69%). Only four studies were conducted in low- and middle-income countries. Interventions evaluated included provider-led interventions, integrated screening and treatment, comprehensive disease management programs, and a quality improvement collaborative. The majority of the studies focused on concordant MM, with hypertension or diabetes as the primary diseases of focus. More than half of the selected studies used simulated Markov models, while only three used individual sampling models. Few studies were explicit about their data validation approaches against local data, the quality of data incorporated in the models, and internal and external consistency.

CONCLUSIONS: There is a need for more studies using individual patient simulation models to evaluate the impact of integrated care for cardiometabolic MM. Furthermore, the shortcomings of reporting of heterogeneity, input data, methodological uncertainty, and validation indicate a need for more transparent reporting.