To Repeat or Not to Repeat: An Evaluation of Critical Values in Chemistry Laboratory Testing

Abstract

Abstract It is common laboratory practice to repeat critical values before reporting the result. However, the benefits of repeat testing are unclear. There are slight differences in values whenever tests are repeated, owing to the inherent imprecision of instruments. Repeating critical values delays reporting of results and, ultimately, clinical intervention. Modern instruments have ultrasensitive level sensors, clot detectors, and better precision, making this practice costly and clinically inefficient. This study aimed to evaluate critical values pairs (original and repeats) to determine if repeats were necessary. Our instruments are programmed to automatically repeat critical values for several analytes. In total, 3,891 critical value pairs (potassium, n = 1111; sodium, n = 677; glucose, n = 987; calcium, n = 214; phosphorus, n = 86; magnesium, n = 399; and bicarbonate/CO2, n = 417) that were performed from April to July 2018 at our core laboratory were analyzed. Data were extracted using software developed to evaluate clinical laboratory data for quality assurance. Analysis of the distribution of critical values within and outside the analytical measurement range (AMR) for each of the analytes was performed. The repeat pairs were analyzed for significant differences (>CAP allowable error), mean absolute and percentage difference between the two determinations, and the number of results that were no longer critical on repeat testing. A cost-time analysis was also performed. For all critical values, a total of 49 (1.26%) showed significant differences between the initial and the repeated results. There were significant differences in 35 (0.92%) of the 3,800 repeats within and in 14 (15.38%) of the 91 repeats outside the AMR. For the 3,800 critical values within the AMR, the mean absolute difference and percent difference between the two testing determinations were: potassium, 0.02 mmol/L (0.42%); sodium, 0.19 mmol/L (0.13%); glucose, 0.44 mg/dL (0.16%); calcium, 0.01 mg/dL (0.14%); phosphorus, 0.01 mg/dL (0.16%); magnesium, 0.01 mg/dL (0.31%); and bicarbonate (CO2), 0.57 mEq/L (1.73%), respectively. A total of 254 (6.53%) initial critical values turned out to be noncritical when repeated, although most were just above or below the cutoff. During the study period, the analytical time for critical value repeats was around 355 hours and the cost was just over $1,300. Data from our study suggest that repeated testing is very similar to the original result and therefore redundant. A higher percentage of results had a significant difference when outside the AMR and repeating just these results can be considered. This study did not capture results that were just below or above the critical value cutoffs, and several of these would have likely been critical on repeat testing. Our analysis suggests that these chemistry critical values should not be routinely repeated, allowing for potential cost savings and improved patient care by reporting critical values sooner, resulting in faster clinical intervention.