Evaluation of i-STAT Creatinine Assay

Abstract

Background: We evaluated a new, 2-min blood creatinine method using the hand-held i-STAT analyzer. Good results have already been reported using this analyzer for 10 methods including electrolytes, TCO₂, pH, PCO₂, bicarbonate, glucose, hemoglobin and urea for uremic blood, hemodialysate and peritoneal effluent. Methods: Evaluation included study of imprecision and accuracy. Results: Imprecision studies gave excellent results, including those for reproducibility of 6 solutions with a mean of 10 repeats and coefficients of variation (CVs) of 0.4–3.4%, and also the mean of the differences between 33 duplicate blood specimens which was 2.2% of the specimen mean. To assess accuracy, we compared results of 149 tests by i-STAT and Beckman Synchron CX7 methods. The difference between the two means was 2.6% and the mean of all differences was 10.9% with i-STAT results higher, especially when blood creatinine values were <100 µmol/l (1.1 mg/dl) indicating the need for a slightly higher upper limit of the normal range. The correlation coefficient between the two methods was 0.99, the slope 1.0 and the intercept –5.0 µmol/l (–0.06 mg/dl). We assessed the recommended creatinine correction for variation in PCO₂above and below 40 mm Hg, but our results did not suggest the need for such a correction in our range of 27–64 mm Hg; omission would remove a major method disadvantage. Assays of hemodialysate and peritoneal effluent were also satisfactory. Conclusions: The i-STAT creatinine method is simple and rapid and our evaluation showed satisfactory accuracy and precision. However, results were on average slightly higher than for the Beckman Synchron CX7 method.