Influence of care delivery models on quality of diabetes care among individuals with schizophrenia

Objective: Managing chronic conditions in individuals with severe mental illnesses is critical for amending health disparities in this vulnerable group. The study aimed to compare the management and outcomes of diabetes care under different care models in individuals with schizophrenia in Taiwan.

Methods: A population-based retrospective cohort comprising incident cases of diabetes in individuals (N = 9,109) with schizophrenia (ICD-9-CM code 295) in Taiwan between 2008 and 2015 was selected using the National Health Insurance Research Database. Generalized estimating equation (GEE) modeling was used to compare 3 care models: the sole-physician model, the colocation model, and the different-facilities model. Each individual was followed up for 3 years. Propensity score matching was used to address potential selection bias.

Results: Patients in the sole-physician model had the highest number of recommended routine examinations (incident rate ratio [IRR] = 1.2; 95% CI, 1.1–1.2) and the highest likelihood of having regular diabetes-related visits as recommended (odds ratio [OR] = 2.6; 95% CI, 2.1–3.2), followed by those in the colocation model (number of recommended routine examinations: IRR = 1.1; 95% CI, 1.1–1.2; likelihood of regular visits: OR = 1.6; 95% CI, 1.3–1.9) and those in the different-facilities model. However, the sole-physician group had a significantly higher likelihood of admission for diabetes-related ambulatory care sensitive conditions within 1 year (OR = 1.9; 95% CI, 1.3–2.8) and 3 years (OR = 1.6; 95% CI, 1.2–2.1) than its counterparts. Within the sole-physician group, patients of psychiatrists had more favorable disease outcomes than those of non-psychiatrists.

Conclusions: The sole-physician and colocation models may significantly improve the process quality of diabetes care; however, such models alone are not sufficient to improve diabetes outcomes.