The exploration of Guardian real-time continuous monitoring system in high-risk infants with hypoglycemia

Ning Liu,Guoqiang Cheng,Peng Zhang,Kai Yan,Wenhao Zhou

DOI: https://doi.org/10.3760/cma.j.issn.2095-428X.2019.01.005

2019-01-01

Abstract:Objective To explore the feasibility of Guardian real-time continuous monitoring system (GRT) in high-risk infants with hypoglycemia.Methods The glucose of 72 infants admitted to Neonatal Intensive Care Unit (NICU) of Children's Hospital of Fudan University between May 2015 and December 2016 were detected by GRT (provided by Medtronic),there were 43 males and 29 females with gestational age [(35.25 ± 4.45) weeks],and birth weight [(2 385.69 ± 1 062.63) g].At the same time,the capillary glucose was monitored intermittently detected by using ACCU-CHEK(R).Results A total of 1 134 paired glucose levels were collected.A good correlation between the paired capillary[(5.23 ± 1.96) mmol/L] and GRT continuous monitor measurements [(5.19 ± 1.99) mmol/L] was found(r =0.88,P ＜ O.05),and the value of mean absolute difference (MAD) was 11.88％.Through the Clarke error grid analysis,there were 98.24 ％ points located in region A and region B,only 20 paired(1.76％) glucose located in region D.Hypoglycemia was defined as glucose concentration ＜ 2.6 mmol/L,while hyperglycemia was defined as glucose concentration ＞ 7.0 mmol/L.A high/low blood glucose detected by peripheral blood glucose monitoring is a high/low blood glucose event,and a high/low blood glucose detected by GRT in a continuous period is a high/low blood glucose event.Twelve cases of hypoglycemia,26 episodes of hypoglycemia,29 cases of hyperglycemia,and 151 episodes of hyperglycemia were detected by using fast blood glucose meter.Twenty-six cases of hypoglycemia,88 episodes of hypoglycemia,38 cases of hyperglycemia,and 229 episodes of hyperglycemia were detected by using GRT.Sixty-two episodes of hypoglycemia were not detected by the fast blood glucose meter,with 14 episodes lasting longer than 30 min,4longer than 60 min,and 5 longer than 90 min.There were 78 hyperglycemic episodes that were not detected by the fast blood glucose meter,with 44 episodes lasting longer than 30 min,15 longer than 60 min,and 70 longer than 90 min.There were significant differences in the testing of hypoglycemia and pathoglycemia(x2 =7.00,18.60;all P ＜ 0.05),but,there was no significant difference in the testing of hyperglycemia between the 2 kinds of detection(x2 =2.26,P ＞0.05).GRT was acceptable to parents,medical staffs and children as only 3 infants bled when the continuous glucose sensor was imbedded,but there was no progressive increase in blood loss.There was no redness,edema,infections,or effusions at the insertion sites.Conclusions The use of GRT continuous monitoring system in high-risk infants of hypoglycemia was feasible.Compared with the conventional intermittent peripheral blood glucose monitoring,more abnormal blood glucose events and their duration could be detected.