The hemolysis index as a tool for monitoring mild hemolysis in biochemical assays at the emergency laboratory

Hemolysis is a common pre-analytical error that can affect the accuracy of biochemical assay results. The aim of the study is to investigate the impact of hemolysis on the measurement of nine emergency biochemical parameters: glucose, creatinine, calcium, potassium, sodium, Aspartate Aminotransferase (AST), Alanine Aminotransferase (ALT),total bilirubin, and direct bilirubin. The study included blood samples collected from various units of the maternity hospital, including high-risk pregnancies follow-up, postpartum care, gynecology, neonatology, and intensive care. The study employs a Hemolysis Index (HI) to monitor and assess sample quality, determining their suitability for the analysis of the nine aforementioned biochemical parameters. The results reveal that hemolysis can significantly affect the accuracy of tests results for the studied parameters, particularly for potassium (increased 4-12 times) and AST (increased 1.47-48.55 times). Three of the studied substances showed a significant, negative, and moderate correlation with HI after the osmotic shock was induced, including AST (r = -0.501, p = 0.006), ALT (r = -0.516, p = 0.004), and sodium (r = -0.598, p = 0.001). Conversely, no significant association was found for the other parameters, which are: glucose (r = 0.079, p = 0.639); creatinine (r = 0.140, p = 0.402); direct bilirubin (r = 0.292, p = 0.075); total bilirubin (r = 0.272, p = 0.114); calcium (r = 0.215, p = 0.196); and potassium (r = -0.188, p = 0.258). Our findings indicate that HI may not be helpful for calculating a predicted value for samples with HI ranging between 87.30 and 295.9. Therefore, it is crucial to establish a threshold for the degree of hemolysis beyond which releasing the result would be considered potentially harmful to the patient especially new-borns and premature infants. In conclusion, the appropriate use of HI in clinical laboratories can enhance patient care quality by minimizing the risk of misdiagnosis leading to inappropriate treatment, particularly when requesting a second sample is not feasible.