Performance of a hybrid CGM protocol in a renal/general medicine population

In 2024, we published a hybrid continuous glucose monitor (CGM)-based glucose monitoring protocol which aimed to replace 75% of point of care (POC) glucose tests for treatment decisions on inpatients primarily admitted with infectious indications. Further performance evaluation in general medicine and renally compromised patients was of additional interest. CGM accuracy was estimated via error grid analysis and mean absolute relative difference (MARD) calculation between CGM-POC glucose pairs. Hypoglycemic events and the number of POC glucose tests avoided were collected. Fifty-four patients were evaluated across 62 inpatient visits. The overall MARD between 732 paired CGM and POC glucose values was 11.9 (±10.8)%; after protocol-guided removal of 258 paired values, the MARD was 10.0 (± 8.1)%. Twenty-four hypoglycemic (<70 mg/dL) events were identified. CGM replaced 53% of POC glucose tests. CGM-guided insulin therapy has potential as an acceptable replacement for POC glucose in the context of inpatient noncritically ill diabetes management.