Algorithm for Differential Diagnosis of in Vitro and in Vivo Hemolysis in Anticoagulated Blood Specimens

ABSTRACT

BACKGROUND: Hemolytic specimens contain components that interfere with clinical laboratory results. We evaluated previously published hemolysis indices (HI) and developed an algorithm for differentiating between mechanical hemolysis and immune-mediated hemolysis based on complete blood count (CBC). METHODS: Sixty-three residual EDTA (ethylenediamine tetraacetic acid)-anticoagulated blood specimens were obtained during regular health check-ups, and each specimen was divided into 3 aliquots (A control, B, and C group). Aliquots B and C were mechanically hemolysed by 2 and 5 aspirations, respectively, using a 25-gauge needle before testing; aliquot A was analysed immediately without hemolysis. Additionally, we collected 36 specimens from patients suspected of having immune-mediated hemolysis after thorough reviewing their various laboratory results including direct antiglobulin test. We compared CBC parameters between the groups (A, B, C, D [B+C], and E [immune-mediated hemolysis group]). RESULTS: Our HI scoring system using the sum of red blood cell ghosts, measured hemoglobin-calculated hemoglobin, mean corpuscular hemoglobin concentration-corpuscular hemoglobin concentration mean, and mean platelet volume rather than mean corpuscular hemoglobin, effectively identified mechanical hemolysis; the results were similar to those of previous studies. Furthermore, the HI score using the sum of mean corpuscular volume, red cell distribution width, hemoglobin distribution width, polymorphonuclear %, and neutrophil % differentiated mechanical hemolysis from immune-mediated hemolysis (cut-off, 9; sensitivity, 91.7%; specificity, 92.9%; area under the receiver operating characteristic curve, 0.965 [95% confidence interval, 0.924–0.988]). CONCLUSIONS: The newly developed algorithm may provide effective screening for detecting hemolysis and differential diagnosis of mechanical hemolysis and immune-mediated hemolysis based on CBC results.