38-OR: Implementation of an EHR-Integrated Hospital-Wide CGM Protocol for Insulin Dosing

Objective: The study aimed to develop and assess a standardized electronic health record (EHR)-integrated hospital-wide protocol for personal continuous glucose monitor (CGM) use and CGM-based insulin dosing. The focus was on feasibility and acceptability from nurses and patients.

Methods: At Stanford Health Care, a hospital-wide protocol was implemented to support inpatient use of home CGM devices in lieu of frequent fingerstick blood glucose (FSBG) monitoring for insulin dosing. It includes requirements specific to each CGM device and protocols for ongoing accuracy monitoring through intermittent CGM-to-FSBG comparisons (“validations”) integrated directly into the EHR. User experience surveys were conducted among a subset of patients and nurses who used the protocol.

Results: From Nov 2022-Aug 2023, 135 patients used a personal CGM under the protocol in 185 inpatient encounters. This included 26% non-English speakers and 27% with type 1 diabetes. Most used CGMs were Dexcom G6 (44%), Libre 2 (42%), and Libre 3 (5%). Automated insulin delivery was used in 23% of encounters. Of 1506 CGM validations, 87.7% met the requisite accuracy criteria (“%20/20”) for CGM-based insulin dosing, and 99.3% fell within Clarke zones A or B. Surveys were completed by 27 nurses and 46 patients. A majority of nurses found glucose management under the protocol effective (74%), easy to use (67%), and efficient (63%); 80% of nurses preferred inpatient CGM to FSBG. A majority of patients liked glucose management under the CGM protocol (63%), reported positive interaction with nursing staff about CGM use (63%), and felt the number of interruptions related to diabetes management was reasonable (63%).

Conclusion: The inpatient CGM protocol, fully EHR-integrated and with real-time device accuracy monitoring, allowed for insulin dosing based on personal CGM data. It received favorable feedback from both nurses and patients, and is the first to offer guidance for multiple CGM types, tailored individually according to CGM type.