Exact and quick measurements of basic laboratory parameters are important in selected patients in the perioperative period. Depending on the capabilities of a hospital's central laboratory, the anaesthesiologist may only obtain such laboratory tests after unacceptable delays. This problem may be overcome by a new bedside measurement device that has become available from i-STAT Corporation, Princeton, USA. The hand-held, battery-driven analyser accepts blood specimens that are injected into a disposable cartridge (EG7+) and measures acidity, blood gas tensions, haematocrit, and electrolytes. The aim of this study was to determine the accuracy of such measurements by comparing them with measurements obtained by conventional laboratory test methods.
Methods: Heparinised arterial blood specimens were collected in duplicate from 49 surgical patients. Measurements of ionised calcium (Ca), sodium (Na), potassium (K), pH, pCO2, pO2, base excess (BE), haematocrit (Hct), and haemoglobin (Hb) obtained by the i-STAT analyser were compared with measurements from the calibrated analysers ABL 615 and EML 100 (Radiometer, Copenhagen). Because the i-STAT analyser calculates the Hb concentration from a conductometrically measured Hct, 19 blood specimens were centrifuged in order to compare test results with conventionally obtained Hct and Hb values. As the Hct test sensitivity with the i-STAT changes with diluted blood due to its low albumin concentration, Hct and Hb measurements during cardio-pulmonary bypass (CPB) must be corrected by activating an analyser-implemented correction algorithm (Hct/CPB and Hb/CPB). Correlation analysis was performed between conventional measurements and i-STAT values (Ca, Na, K, Hct, pCO2, pO2), between values that the i-STAT analyser derives (Hb, HCO3, BE) and conventionally obtained results, and between normal and CPB-corrected Hct and Hb values. Accuracy was judged according to the national quality standard, which requires test results to lie within the 95% confidence interval of conventional tests.
Results: Each blood specimen was analysed: erroneous results or technical failures did not occur. Measurement of one set of i-STAT values required 2.5 min. Correlation coefficients (r) between conventional and i-STAT results were: 0.85 for CA, 1.0 for K; 0.86 for Na; 0.99 for pH; 0.98 for pCO2; 0.99 for pO2; 0.93 for HCO3; 0.93 for BE; 0.46 for Hb values not corrected for CPB and 0.95 for CPB-corrected Hb; and 0.74 for Hct values not corrected for CPB and 0.98 for CPB-corrected Hct. The correlation coefficient for Hct between centrifuged and CPB-uncorrected i-STAT values was 0.81 and that for CPB-corrected values was 0.98. National accuracy requirements were not met for tests of: Ca (by 0.02 mmol/l); pH (by 0.01); pO2 including hyperoxic values (by 26.7 mmHg, but were met for pO2 values < 200 mmHg); Hb (by 1.6 g/dl); Hb/CPB (by 0.8 g/dl); and Hct (by 6.5%, but were met for Hct/CPB values). All other tests fulfilled the required standards.
Conclusion: This analyser is easy to use, reliable, and portable, and therefore suitable for the operating room, for analyses during emergencies, on peripheral wards, for preclinical screening, or at times when availability of lab tests is time-consuming or limited. The test accuracy for electrolytes, blood gases, and Hb is high enough to justify routine use of the i-STAT analyser in clinical practice. That the nationally required quality standards for Ca, pH, and Hb were not met is not of importance because the measured deviation was too small to have clinical relevance. When analysing diluted blood with a low Hct and low oncotic pressure, it is important to activate the analyser's correction algorithm "CPB", because the obtained results will then comply with the required accuracy.