Impact of Retrospective Calibration Algorithms on Hypoglycaemia Detection in Newborn Infants Using Continuous Glucose Monitoring

BACKGROUND: Neonatal hypoglycaemia is common and may cause serious brain injury. Diagnosis is by blood glucose (BG) measurements, often taken several hours apart. Continuous glucose monitoring (CGM) could improve hypoglycaemia detection, while reducing the number of BG measurements. Calibration algorithms convert sensor signals into CGM output. Thus, these algorithms directly affect measures used to quantify hypoglycaemia. AIM: Quantify the effects of recalibration and filtering of CGM data on measures of hypoglycaemia (BG < 2.6mmol/L) in neonates. METHODS: CGM data from 50 infants were recalibrated using an algorithm that explicitly recognised the high-accuracy BG measurements available in this study. CGM data were analysed as: 1) Original CGM output; 2) re-calibrated CGM output; 3) re-calibrated CGM output with post-calibration median filtering; and, 4) re-calibrated CGM output with pre-calibration median filtering. Hypoglycaemia was classified by: number, duration, severity and hypoglycaemic index. RESULTS: Recalibration increased the number of hypoglycaemic events (161 to 193), hypoglycaemia duration (2.2% to 2.6%) and hypoglycaemic index (4.9 to 7.1μmol/L). Median filtering postrecalibration reduced hypoglycaemic events from 193 to 131, with little change in duration (2.6% to 2.5%) and hypoglycaemic index (7.1 to 6.9μmol/L). Median filtering pre-recalibration resulted in 146 hypoglycaemic events, a total duration of hypoglycaemia of 2.6%, and a hypoglycaemic index of 6.8μmol/L. CONCLUSIONS: Hypoglycaemia metrics, especially counting events, are heavily dependent on CGM calibration BG error and the calibration algorithm. CGMs tended to read high at lower levels, so when high accuracy calibration measurements are available it may be more appropriate to recalibrate the data.