Impact of hemolysis on coagulation testing: comparative analysis of two different analyzers using photometric and mechanical clot detection techniques

BACKGROUND: Analytical methods for clot detection may differ in susceptibility to hemolysis, especially photometric versus mechanical systems. METHODS: Two coagulation analyzers have been compared: Werfen ACL TOP 750, photometric clot detection versus Stago sthemO 301, mechanical clot detection, using 100 routine plasma samples (50 hemolyzed, 50 non-hemolyzed). Prothrombin time (PT), activated partial thromboplastin time (APTT) and fibrinogen were measured in duplicate per analyzer. Analytical imprecision (coefficient of variation; CV%), correlation (Spearman’s rank) and linear fit were calculated, while differences were evaluated using the Mann-Whitney U Test. Bootstrap and hypothesis tests for equality of regression slopes and intercepts were used to compare inter-analyzer relationships in non-hemolyzed and hemolyzed plasma samples. RESULTS: PT, APTT and fibrinogen values did not significantly differ between non-hemolyzed and hemolyzed samples (all p>0.05). Correlation between analyzers was strong across all parameters (all r>0.74). Bootstrap test revealed no significant differences of inter-analyzer correlations between non-hemolyzed and hemolyzed samples for all parameters (all p>0.05). The hypothesis tests for equality of regression slopes and intercepts showed significant differences between non-hemolyzed and hemolyzed samples for fibrinogen but not for PT and APTT. Imprecision was higher with ACL TOP, especially for APTT in hemolyzed samples (CV% 1.6 versus 0.6%, p<0.001). Fibrinogen measurements on ACL TOP displayed approximately double CV% compared to Stago sthemO, irrespective of hemolysis. CONCLUSIONS: Both analyzers showed comparable performance for PT and APTT in hemolyzed and non-hemolyzed samples, while fibrinogen displayed a different trend. Mechanical clot detection provided better overall precision for fibrinogen and for APTT in hemolyzed specimens.