Supplemental Material: Continuous Glucose Monitoring-Guided Intravenous Insulin Infusion in Hospitalized Patients With Diabetes Emergencies: A Real-World Clinical Experience

Aims: Glucose management in hospitalized patients with severe hyperglycaemia needing intravenous continuous insulin infusion (CIII) is complex. Real-time continuous glucose monitoring systems (rtCGM) provide a more comprehensive glucose assessment compared to point of care testing (POC). We hypothesise that rtCGM compared to POC improves inpatient diabetes care in patients with hyperglycemic emergencies. Methods: In this single-centre retrospective study, data from 225 patients admitted due to hyperglycemic emergencies requiring CIII on a specialised non-ICU diabetes ward were analysed. Data were compared between patients with rtCGM (122 patients) and POC (103 patients). After stabilisation of glucose parameters patients were switched to subcutaneous insulin application. Performance and outcome parameters included mean glucose concentrations, glucose variability, frequency and degree of hypoglycemic exacerbations, accuracy for subcutaneous insulin dosing and length of stay (LOS). Results: Demographic and diabetes-specific characteristics did not differ between the two patient groups. In patients with rtCGM mean blood glucose concentrations were significantly lower during hospitalization (167.7+28.6 vs 180.5+31.2 mg/dl; p<.01) as was the incidence of hypoglycemic events (74 vs 193, p<.01). Transition from CIII to s.c. insulin injection was facilitated in patients with rtCGM and required insulin doses were lower at discharge (11.8 IU with rtCGM vs. vs 16.2 IU with POC, p.03). Moreover, length of stay was significantly shorter in patients with rtCGM compared to POC monitoring (7.9+2.9 vs. 9.2+3.5; p<.01). Conclusion: CGM use in inpatients with hyperglycemic emergencies receiving CIII was associated with a reduced LOS probably due to a more accurate diabetes management and compared to conventional POC.