CGM works in select ICU patients
A new study reports continuous glucose monitors gave reliable blood‑sugar readings in select ICU patient groups, supporting broader inpatient CGM use for more responsive glycemic management Medical Dialogues. That could change insulin protocols and monitoring workflows in critical care.
The pilot placed Dexcom G7 sensors on 16 consented ICU patients over a 9‑month period reported). Investigators collected 941 near‑simultaneous CGM vs routine care pairs, reporting a median sensor duration of 5.5 days and an overall MARD of 12.25% with 99.7% of values in Clarke Error Grid zones A or B. reported) No device‑related adverse events were recorded, and qualitative interviews showed nursing staff reported no concerns about adding CGM to routine ICU workflow during the study period noted). The same report warned accuracy may decline in patients on high‑dose vasopressors observed); a separate 2024 ICU cohort found suboptimal CGM performance during higher‑dose vasopressor infusions and within the first 36 hours after sensor insertion found). The study’s MARD (12.25%) aligns with prior inpatient research that has reported lab‑CGM MARDs near 11.4% and POC‑CGM MARDs up to ~15–17% in real‑world hospital cohorts summarized). Larger validation efforts are already registered: a University of Massachusetts trial of the Dexcom G7 in the ICU ran from Feb 20, 2024 to Jul 25, 2025 (clinical trial record NCT06190808), and other multicenter studies comparing CGM‑guided insulin protocols to point‑of‑care regimens are ongoing listed). A recent retrospective analysis of CGM‑guided insulin infusion protocols reported improved glycemic metrics versus standard point‑of‑care titration, signaling the specific workflow changes ICUs are studying if larger trials confirm CGM accuracy and safety in targeted patient groups reported).
