Accuracy and reliability of continuous blood glucose monitor in post-surgical patients K. YAMASHITA 1 , T. OKABAYASHI 2 , T. YOKOYAMA 1 , T. YATABE 1 , H. MAEDA 2 , M. MANABE 1 and K. HANAZAKI 2 Departments of 1 Anesthesiology and Critical Care Medicine and 2 First Department of Surgery, Kochi Medical School, Kochi, Japan Background: The STG-22t is the only continuous blood glucose monitoring system currently available. The aim of this study is to determine the accuracy and reliability of the STG-22t for continuously monitoring blood glucose level in post-surgical patients. Methods: Fifty patients scheduled for routine surgery were studied in surgical intensive care unit (ICU) of a university hospital. After admission to the ICU, the STG- 22t was connected to the patients. An attending physician obtained blood samples from a radial arterial catheter. Blood glucose level was measured using the AB- Lt800FLEX immediately after blood collection at 0, 4, 8, and 16 h post-admission to the ICU (total of 200 blood glucose values). Results: The correlation coefficient (R 2 ) was 0.96. In the Clarke error grid, 100% of the paired measurements were in the clinically acceptable zone A and B. The Bland and Altman analysis showed that bias  limits of agreement (percent error) were 0.04(0.7)  0.35(6.3) mmol (mg/dl) (7%),  0.11(  2)  1.22(22) (15%) and  0.33(  6)  1.28(23) (10%) in hypoglycemia (o70(3.89) mmol (mg/dl), normoglycemia (3.89(70)–10(180) mmol (mg/dl), and hyperglycemia (410(180) mmol (mg/dl), respectively. Conclusions: The STG-22t can be used for measuring blood glucose level continuously and measurement results are consistent with intermittent measurement (percentage error within 15%). Therefore, the STG-22t is a useful device for monitoring in blood glucose level in the ICU for 16 h. Accepted for publication 20 August 2008 r 2008 The Authors Journal compilation r 2008 The Acta Anaesthesiologica Scandinavica Foundation H YPERGLYCEMIA often occurs in critically ill pa- tients 1,2 and has also been associated with adverse outcomes in surgical patients due to increased risk of infection and slower wound heal- ing. 3 Recent research suggests that tight control of glucose might reduce morbidity and mortality in critically ill patients. 1,4–6 Blood glucose level has been shown to exceed 10(180) mmol (mg/dl) in about 45% of intensive care unit (ICU) patients despite frequent blood glucose evaluation. 7 The insulin infusion protocol used for monitoring development of normoglycemia is usually complex, and thus, may not be precisely followed by busy ICU staff. 8–10 A continuous blood glucose monitor would be beneficial for maintaining target blood glucose levels in ICU patients. Two types of continuous glucose monitoring systems are currently available: a continuous sub- cutaneous glucose monitor 11,12 and a continuous blood glucose monitor. The continuous subcuta- neous glucose monitoring system is less invasive than a continuous intravenous blood monitoring system with regard to bleeding, infection, throm- bus formation, and nerve damage, among others. Although the continuous subcutaneous glucose monitor is less invasive, Chase et al. 13 reported that it has a larger margin of error than pin-prick methods for measuring blood glucose. The STG-22t (Nikkiso Inc., Tokyo, Japan) is currently the only continuous blood glucose mon- itoring device in the world. Originally, the STG-22t was used as an artificial endocrine pancreas. It has a closed-loop glycemic control system enabling continuous intravenous blood glucose monitoring through a glucose sensor electrode and maintains appropriate blood glucose levels through auto- matic insulin and glucose infusions. 14 The STG- 22t has also been used for evaluating insulin resistance in patients with insulinomas. 15 Given its previous uses, the system has potential for monitoring blood glucose in critically ill patients, although its accuracy and reliability in this popula- tion have not been fully investigated to date. In a previous report, the STG-22t was identified as a 66 Acta Anaesthesiol Scand 2009; 53: 66–71 Printed in Singapore. All rights reserved r 2008 The Authors Journal compilation r 2008 The Acta Anaesthesiologica Scandinavica Foundation ACTA ANAESTHESIOLOGICA SCANDINAVICA doi: 10.1111/j.1399-6576.2008.01799.x