Evaluation of point-of-care testing in critically unwell patients: comparison with clinical laboratory analysers and applicability to patients with Ebolavirus infection.

Data on the performance of point-of-care (POC) or near-patient devices in the management of critically unwell patients are limited, meaning that there are demands for confirming POC test results in the routine clinical laboratory and so potentially leading to delay in treatment provision. We evaluated the performance of the i-STAT CHEM 8+ and CG4+, Hemochron Signature Elite, HemoCue Hb 201+ and WBC Diff Systems on whole blood collected from medical and surgical patients admitted to the intensive care unit at an Australian tertiary care hospital. Measurements obtained for haematology, coagulation, biochemistry and arterial blood gas parameters using POC devices were compared against clinical laboratory analysers (XE-5000, STA-R Evolution, Dimension Vista 1500 and ABL800 FLEX). Bland-Altman and Passing-Bablok regression plots were constructed to assess agreement. Good correlation was defined as a bias of <10% between the POC device and the reference method. Forty arterial blood samples were collected from 28 patients. There was good correlation demonstrated for sodium, potassium, chloride, ionised calcium, glucose, urea, haemoglobin and haematocrit values (i-STAT Chem 8+); pH, pCO(2), bicarbonate and oxygen saturation (i-STAT CG4+); haemoglobin, white cell, neutrophil count and lymphocyte counts (Hemocue); and internationalised normal ratio (INR; Hemochron Signature Elite), but not creatinine, anion gap, pO(2), base excess, lactate, eosinophil count, prothrombin and activated partial thromboplastin time. POC devices were comparable to clinical laboratory analysers in measuring the majority of haematology, biochemistry and coagulation parameters in critically unwell patients, including those with infections. These devices may be deployed at the bedside to allow 'real-time' testing to improve patient care.