Abstract
Background: Effective treatment and close monitoring in children with congenital hyperinsulinism (HI) are important to prevent hypoglycemic-associated brain damage. The current monitoring approach involves measuring plasma glucose intermittently, but this does not provide a comprehensive assessment of glycemic control and may fail to detect episodes of hypoglycemia. Objective: To determine whether Dexcom G5®, a continuous glucose monitoring system (CGMS), is an accurate and effective method for monitoring glycemic control in children with HI. Methods: Cross-sectional, observational study in 15 children with HI. Participants wore a blinded Dexcom G5® device for 2 weeks. At the end of 2 weeks, data from the Dexcom G5® and home glucose meter were downloaded and analyzed. Results: Fourteen children (15–67 months) completed the study. Using Bland-Altman analysis, the mean (SD) difference between 1,155 paired CGM and glucose meter readings was –8.09 (53.76). The sensitivity and specificity of CGM to detect hypoglycemia (<70 mg/dL) were 86 and 81.4%, respectively. The positive predictive values for hypoglycemia and severe hypoglycemia (<54 mg/dL) detected by CGM were low (50.3 and 14.8%, respectively), while the negative predictive values were high (96.4% for glucose <70 mg/dL and 99.1% for glucose <54 mg/dL). Conclusion: Our study showed that CGM is not a reliable method to monitor for hypoglycemia, given the high number of false positive hypoglycemia readings. However, CGM can be useful in preventing unnecessary checks by glucose meter during times of normoglycemia. Therefore, the benefits of using CGM in patients with HI would be in guiding the need to check plasma glucose by glucose meter rather than point accuracy.
