Affiliation:
1. Department of Community Medicine and School of Public Health, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India
Abstract
Background:
Health-care professionals routinely measure blood glucose of all critically ill patients in intensive care unit (ICU). The traditional reference standard, venous blood glucose requires venipuncture, and laboratories usually return the test results after 60 min. We wondered if a point of care (POC) capillary glucose measured by glucometer could be as accurate and reliable as blood venous glucose.
Methods:
We studied consecutive patients admitted to an ICU between May 1, 2009, and May 10, 2009. We compared the diagnostic accuracy of two-POC glucometers with laboratory venous glucose, the reference standard. We used Bland–Altman plots and the Clark error grid method to analyze the results.
Results:
This study included a total of 110 patients 38 (34%) women; mean age 52.1 years (standard deviation [SD] =17.3); range = 14–85 years. Fourteen patients (12%) were known to have diabetes. The mean glucose value (glucometer 1) was 152.9 mg/dL (SD = 83.1); range = 48–501 mg/dL; that by glucometer 2 was 152.2 mg/dL (SD = 76.2); range = 30–458 mg/dL and by the laboratory was 148.6 mg/dL (SD = 81.5); range = 52–480 mg/dL. Of the 110 subjects, 2 (2%) had blood glucose below 70 mg/dL; 85 (77%) between 70 and 180 mg/dL; and 23 (21%) had blood glucose exceeding 180 mg/dL. The Bland–Altman plot showed a bias of 4 mg% (95% confidence intervals = −9.8–+1.1); and the limits of agreement were −63 and +54 mg%. The area under the receiver operating characteristic curve for the two glucometers was 0.92 and 0.93, respectively. The error grid analysis showed that 80% (78/110) of samples lay within Zone A, 22 (110) in Zone B, and 1% (1/110) in Zone C.
Conclusions:
POC glucose, measured by glucometers was in agreement with the venous glucose estimation. Both glucometers were equally accurate and performed uniformly well across the wide range of blood glucose values.