Letter: Beyond Accuracy: Clinical Considerations for Personal Continuous Glucose Monitoring Use During Hospitalization

To the Editor,

We read with great interest the recent publication titled “Real-World Performance of Personal Continuous Glucose Monitors During Hospitalization” by Dumitrascu et al. ¹ The authors offer important insights into the accuracy and clinical utility of outpatient-inserted continuous glucose monitors (CGMs) during noncritical hospital admissions. Their findings provide valuable support for the broader integration of CGM technology in inpatient settings. Building on their important work, we would like to offer additional clinical perspectives that may further enrich the ongoing discussion.

The study convincingly demonstrated that CGM accuracy was acceptable for glucose values above the hypoglycemia range. However, from a clinical management standpoint, a nuanced appreciation of CGM performance during dynamic glycemic shifts is particularly critical. Hospitalized patients often experience rapid glucose fluctuations—during perioperative periods, corticosteroid administration, or nutritional interventions such as enteral or parenteral feeding.^2,3 These rapid changes may disproportionately affect CGM accuracy due to inherent physiological lag in interstitial glucose measurements. Although the authors appropriately paired CGM and point-of-care (POC) measurements within a 5-min window, a targeted subgroup analysis focusing specifically on periods of rapid glycemic excursions (e.g., postprandial spikes, insulin corrections, or treatment of hypoglycemia) would offer clinically actionable information. Identifying contexts where CGM lag becomes clinically significant could inform protocols for confirmatory POC testing and alert settings during inpatient care.

Moreover, while the study found no correlation between CGM accuracy and static laboratory parameters such as hemoglobin, creatinine, or oxygen saturation, additional sources of variability warrant attention. One such factor is the insulin management approach. The original report did not specify whether insulin dosing followed a standardized protocol—such as a basal–bolus regimen with predefined correction scales—or if dosing strategies varied by provider. In a single-center setting, even institution-wide guidelines may be inconsistently applied, leading to heterogeneity in glycemic control and confounding assessments of CGM accuracy. Another important consideration is device-related variability. The study included both Dexcom and FreeStyle Libre sensors, which differ in calibration requirements, data acquisition intervals, and lag characteristics. Recent evidence has shown substantial differences in CGM metrics between devices used simultaneously in the same patients. ⁴ Without stratified reporting or sensor-specific analysis, this heterogeneity may further complicate the interpretation of overall CGM performance. Future investigations should aim to implement clearly defined insulin protocols and stratify outcomes by CGM device type to improve the reliability and generalizability of findings.

Finally, the glucometric results reported a mean time in range of 58.7%, which the authors presented as acceptable in hospitalized settings prioritizing hypoglycemia avoidance. While this value aligns with recent inpatient studies such as the TIGHT trial, it also suggests an opportunity to further optimize glycemic control. ⁵ Given that CGMs offer continuous data, future efforts might leverage real-time trend analyses (e.g., rates of glucose rise or fall) rather than static thresholds alone to enhance early clinical intervention. This strategy could be particularly valuable in surgical or high-risk medical populations where even short periods of uncontrolled hyperglycemia have been linked to adverse outcomes.

In conclusion, Dumitrascu et al. provide essential foundational evidence supporting the inpatient use of personal CGMs. To build upon this work, future research would benefit from incorporating standardized insulin protocols, examining CGM performance during dynamic glucose shifts, and leveraging trend-based metrics for proactive management. These enhancements could help refine best practices for CGM-guided inpatient glycemic control and support the safe, effective, and scalable integration of CGM technology into routine hospital care.