Evaluation of Blood Glucose Monitoring System in Screening for Neonatal Hypoglycemia: Tighter Accuracy Standard

Abstract

Neonatal hypoglycemia may cause severe neurological damages; therefore, tight glycemic control is crucial to identify neonate at risk. Previous blood glucose monitoring system (BGMS) failed to perform well in neonates; there are calls for the tightening of accuracy requirements. It remains a need for accurate BGMS for effective bedside diabetes management in neonatal care within a hospital population. A total of 300 neonates were recruited from local hospitals. Accuracy performance of a commercially available BGMS was evaluated against reference instrument in screening for neonatal hypoglycemia, and assessment was made based on the ISO15197:2013 and a tighter standard. At blood glucose level < 47 mg/dl, BGMS assessed met the minimal accuracy requirement of ISO 15197:2013 and tighter standard at 100% and 97.2%, respectively.

In healthy neonates, low blood glucose concentrations are frequently observed, such transitional neonatal hypoglycemia normally resolves before 2 or 3 days of age. Prolonged or recurrent hypoglycemia may cause acute systemic damages leading to abnormal neurological outcome.^1-3 Therefore, tight glycemic control is crucial in neonate to differentiate normal physiological responses from underlying disorders. An “operational” threshold of less than 47 mg/dl has been suggested by cumulative evidences to be associated with adverse neurodevelopmental outcomes.^2,4 This cutoff value serves as a useful guideline for clinicians to take appropriate actions, including retesting and continued monitoring of blood glucose, and providing glucose source in oral feeding or intravenous administration, to aim at maintaining blood glucose levels of at least 47 mg/dl.^2,5 Previous blood glucose monitoring system (BGMS) failed to perform well in neonates; there are calls for the tightening of accuracy requirements.^5,6

A total of 300 neonates were recruited from 5 local hospitals in a multi-center clinical trial. Accuracy performance of a commercially available BGMS was evaluated against reference instrument (YSI 2300 PLUS Glucose Analyzer) in screening for neonatal hypoglycemia, and assessment was made based on the International Organization for Standardization (ISO) 15197:2013 and a tighter standard (±10 mg/dl). The minimum acceptable accuracy is as follows: 95% of the individual glucose readings shall fall within ±15 mg/dl of the reference measurement readings at glucose concentrations <100 mg/dl and within ±15% at glucose concentrations ≥100 mg/dl. Neonatal hypoglycemia is defined as a blood glucose level of < 47 mg/dl, of which 30 mg/dl serves as the subthreshold for types of clinical intervention required. At blood glucose level < 47 mg/dl (neonatal hypoglycemia), BGMS assessed met the minimal accuracy requirement of ISO 15197:2013 and tighter standard at 100% and 97.2%, respectively.

Experts argued the minimum requirement of ±15 mg/dl described in current ISO 15197 standards failed to take into account of the low blood glucose and high hematocrit values seen in newborns. ³ This provides a possible rationale to the poor performance of previous self-testing glucometers in neonatal care.^5,6 A tighter accuracy requirement, as proposed in current study (±10 mg/dl), is thus warranted to address the specificity of neonatal blood for effective diabetes management and glycemic control.

This indicated 2.8% (2 out of 72) subjects may have potential false readings at blood glucose < 47 mg/dl. Such report alters clinicians to continued monitor these 2 subjects to take appropriate actions. Precise locations of these 2 subjects at blood glucose < 47 mg/dl were illustrated in scatter plot analysis (Figure 1), in which 1 subject fell below (false negative) and 1 subject fell above (false positive) the ±10 mg/dl acceptance limits. A great concern is to underestimate the severity of hypoglycemia for the “false negative” subject. The glucometer evaluated herein performed well at low blood glucose readings, even under tighter ISO standards.

Figure 1.

Scatter plot analysis on the difference between glucometer (GM700) and reference instrument (YSI 2300 Plus) within the criteria ±15 mg/dl (red solid line) and ±10 mg/dl (blue dashed line).