ABSTRACT
BACKGROUND: Using middleware solutions, it is possible to implement concentration-dependent analyte-specific hemolysis rejection limits. This makes day-to-day reporting of clinical specimens more efficient and potentially lowers sample rejection rates compared to a "one-size-fits-all" approach (i.e., solely based on a single cutoff provided in the package insert). METHODS: Hemolysis interference studies were performed at multiple analyte concentrations for three frequently ordered tests. For each assay, concentration-dependent hemolysis rejection limits were designed based on the total allowable error (TAE) for the analyte as well as the clinical significance of such incurred inaccuracy at the respective concentrations. In general, the following rationale was used: if the interference exceeds 10% (or package insert cutoffs), a comment is placed on the result. If the interference exceeds the TAE, the result will not be reported. Reduction in specimen rejection rates were estimated by comparing the incurred specimen rejection rates when package inserts' vs concentration-dependent hemolysis interference limits were applied to a data set in our institute during a three-month period. RESULTS: Concentration-dependent analyte-specific hemolysis rejection thresholds were designed for three commonly ordered assays that are especially susceptible to hemolysis interference. It is estimated that these novel thresholds for aspartate aminotransferase (AST), lactate dehydrogenase (LDH), and direct bilirubin (DBIL) reduced specimen rejection rates from 9.3% to 1.3%, 31.4% to 4.8%, and 19.9% to 7.1%, respectively. CONCLUSIONS: Concentration-dependent analyte-specific hemolysis rejection thresholds for three commonly ordered assays can reduce rejection rates without significantly compromising the quality of test results.
