Accuracy of Dexcom One+ in Patients with Diabetes or Stress Hyperglycemia Hospitalized in Cardiac Intensive Care Unit

Continuous glucose monitoring (CGM) systems may assist in glucose management for patients in the cardiac intensive care unit (CICU). We aimed to assess the accuracy, feasibility, and tolerability of Dexcom One+ in comparison with standard blood glucose measurements.

From September 2024 to May 2025, we included patients with known diabetes or hyperglycemia on admission >140 mg/dL who were hospitalized in the CICU for acute coronary syndrome and/or heart failure. Sensors were inserted into the upper arms, and glucose readings were obtained using a dedicated receiver. Glucose levels were measured with the Cobas Pulse glucometer (Roche Diagnostics) at the routine frequency in the CICU as part of standard care. Accuracy was evaluated by the mean absolute relative difference (MARD). Clinical performance was assessed through Consensus Error Grid analysis. The feasibility outcome included the number of early sensor detachments and sensor failures. Safety outcomes encompassed skin reactions.

We obtained 999 CGM-reference glucose paired samples from 48 patients (39 with previously diagnosed type 2 diabetes aged 73.5 ± 9.6 years. The mean HbA1c was 7.1 ± 1.3%. Overall, 725 paired samples were obtained during oxygen therapy, 362 during vasopressor infusion, and 280 during combined oxygen therapy and vasopressor infusion. CGM use duration was 4.0 ± 3.2 days. There were 3 reference readings below 70 mg/dL, 658 within the 70-180 mg/dL range, and 338 above 180 mg/dL. Overall, MARD was 11.6% (95% CI: 10.9-12.2). 93.5% of readings were in Zone A, 6.1% in Zone B, and 0.4% in Zone C. No readings were found in Zone D + E. We observed one mild hematoma at the insertion site, three sensor failures, and three early detachments.

In patients in CICU, the Dexcom One+ system showed acceptable accuracy and could support glucose monitoring.