Frequent measurement of blood glucose among hospitalized patients including those in critically ill areas is routinely practiced. Evidence suggests that tight glycaemic control among hospitalized patients reduced complications such as sepsis and has been shown to reduce length of stay. Most commercially available devices have not been assessed in critically ill patients, furthermore point of care (POC) glucose testing among paediatric population has exhibited a discrepancy greater than 20 mg/dL (1 mmol/L). Recent regulatory agencies voiced concern on the validly of POC deceives use in critically ill patients and the lack of adequate data on its accuracy.
This study examined the use of POC glucose measurement using Precision PXP (Abbott Diagnostics) in the monitoring of glucose in critically ill patients. Patients in intensive care units (adult and paediatric), in cardiac care units and those presenting to the emergency department were included in the study. Concurrent POC glucose measurements from 182 patients from those units were compared with laboratory-based glucose values measured using the Hexokinase methods (Cobas, Roche Diagnostics). Sixty three Patients were considered critically ill with a total of 438 samples analysed from all study patients. Glucose values ranged from 25 mg/dL (1.4 mmol/L) to 594 mg/dL (33.0 mmol/L). Bias ranged from −95 to 36 mg/dL (−5.3 to 2.0 mmol/L). Among critically ill patients glucose levels ranged from 50 mg/dL (2.8 mmol/L) to 190 mg/dL (10.6 mmol/L). Overall percentage bias was less than 6%. Clarke and Consensus diagram showed 98100% of data within area A indicating no impact on clinical action. This study showed no significant impact when using the Abbott PXP glucose meter on critically ill patients as compared with a laboratory-based methodology.