Association between processes of diabetes care and health care utilization in patients with diabetes: Evidence from a nationally representative US sample.

BACKGROUND

This study examined the association between quality of processes of diabetes care in terms of adherence to American Diabetes Association (ADA)-recommended guidelines and health care utilization in patients with diabetes.

METHODS

Adults with diabetes were identified from the pool of five panels of the Medical Expenditure Panel Survey, a nationally representative US sample, between 2012 and 2017. The Diabetes Care Survey was used to determine adherence to the ADA-recommended guidelines for processes of diabetes care if all of the following were performed annually: glycosylated hemoglobin check, foot examination, dilated eye examination, lipid panel, influenza immunization, blood pressure check, and dental examination. Health care utilization in terms of inpatient hospitalization, and emergency department (ED) and outpatient visits were estimated using two-part hurdle models.

RESULTS

An estimated 26.3 million adults with diabetes were derived from the pooled 5-panel data, of which 7.8% met the ADA-recommended guidelines for processes of diabetes care, and adherence rates of individual recommendations were generally below 50%. Overall, adults who adhered to the ADA-recommendations were older, non-Hispanic white, and married nonsmokers with private insurance and higher income. Mean inpatient hospital stays, ED, and outpatient visits between ADA-adherent vs nonadherent patients were 0.98 vs 1.62 (P < .001), 0.36 vs 0.39 (P = .074), and 17.9 vs 12.8 (P < .001), respectively.

CONCLUSIONS

Socioeconomic disadvantage and minority status were linked with nonadherence to the ADA-recommended processes of diabetes care. Adherence to the ADA recommendation was associated with significant reduction in inpatient hospitalization and a trend toward less ED visits. Our findings may apply to the United States and are likely to be different in other parts of the world.