Introduction: For decision-making in the ICU, rapid and accurate analysis of vital laboratory parameters is essential. The industry provides devices which analyse these parameters on a decentralised setting and which are designed for use by non-laboratory personnel. We investigated whether accuracy and handling of a new analyser (Ciba-Corning 865, Chiron Diagnostics, Medfield, USA) are good enough for basing clinical decisions on the measured parameters.
Materials and methods: The Ciba-Corning 865 allows measurement of blood gases, electrolytes, haemoglobin, glucose and lactate by use of photometric, ion-selective, enzymatic and electrochemical sensors in less than 18 microliters of whole blood. In a cardiac surgical intensive-care unit the accuracy of the device was tested by comparison to 61 measurements of quality control reagents, 48 tonometered blood samples and 536 parallel measurements in the clinical laboratory. Besides a 10-minute instruction, the participating personnel had no formal training with the device.
Results: The differences between measurements in quality control reagents and tonometered blood and the expected value were lower than 5%. The comparison with clinical laboratory measurements showed correlation coefficients from 0.94 (sodium) to 0.99 (glucose, lactate). The biases in Bland-Altman analyses were below 5%, the limits of agreement were found to be in a clinically acceptable range for all parameters. During the test period no technical problems occurred with the analyser and good acceptance by the personnel was found.
Conclusions: The measured parameters were accurate enough to be used for therapeutic decisions in acute care medicine. Although it should not be a complete alternative to the clinical laboratory, because of rapid analyses, small sample volumes and easy handling the use of the Ciba-Corning 865 is advantageous for patients and users.