876-P: Accuracy of Continuous Glucose Monitoring Compared with Capillary Blood and Venous Plasma Glucose Measurements in Medical Intensive Care Unit Patients

Abstract

Objective: To assess the accuracy of the subcutaneous continuous glucose monitoring (CGM) sensor by comparing to the capillary blood glucose (CBG) and venous plasma glucose (VPG) measurements in the medical intensive care unit (MICU) patients. Subjects and Methods: In a prospective study in patients in the MICUs, Medtronic Enlite sensor was inserted in the abdominal area. Paired sensor glucose readings with reference glucose values (CBG and VPG) were collected from MICU patients receiving intravenous insulin infusion therapy. The accuracy was assessed using the mean absolute relative difference (MARD), surveillance error grid (SEG) analysis and modified Bland-Altman plot. Results: Twelve patients completed the study (age 69.3 ± 11.6 years; BMI 21.6 ± 2.9 kg/m2; APACHE II score 21.8 ± 6.3 and duration of CGM use 108.7 ± 41.1 hours). A total of 353 paired CGM and CBG glucose readings and 125 paired CGM and VPG readings were included in the analysis. Using CBG as the reference, MARD was 7.2%. The modified Bland-Altman plot showed 95% of the limit of agreement (LoA) was -18.4% to 19.2%. The SEG analysis showed that 100% of paired glucose values were within zones A or B. No clinically significant difference in the accuracy was seen between subgroups of vasopressor use (MARD 6.1% in the vasopressor group vs. 6.8% in the non-vasopressor group, p = NS). Using VPG as the reference, the MARD was 8.8%. The modified Bland-Altman plot showed 95% LoA of - 22.6% to 28.2%. The SEG analysis showed 95.2% of glucose readings were within zones A or B. Conclusion: The accuracy of the subcutaneous CGM sensor in MICU patients was comparable to that observed in non-critical care settings, with no device-related adverse events. The sensor accuracy remained within the similar range when using VPG as the references compared to using CBG as the references. No differences in terms of accuracy between the vasopressor and non-vasopressor groups were found in this study. Disclosure N. Laichuthai: None. P. Buranasupkajorn: None. W. Khovidhunkit: Speaker’s Bureau; Self; Amgen, AstraZeneca. Funding Quality Improvement Fund; King Chulalongkorn Memorial Hospital; Thai Red Cross Society (1-62-30101-A-11)