Abstract Objectives Haemolysis is the leading preanalytical cause of sample rejection in coagulation testing. This study prospectively evaluated the impact of spontaneous haemolysis on routine haemostasis assays by comparing paired haemolysed (H) and non-haemolysed (NH) clinical samples. Additionally, analyser flagging behaviour was characterised using an artificially haemolysed sample series. Methods Thirty nine paired of H and NH samples were analysed for prothrombin time (PT), activated partial thromboplastin time (aPTT), and D-dimer (DD) on the Sysmex CN-3000. Mean biases were compared against predefined critical differences (CD) derived from biological variation data. Analyser flagging was additionally assessed using five sample pools across five haemolysis concentration levels (H1–H5). Results When comparing H samples with NH samples, the mean bias was +0.07 s (+0.14 %) for PT (p=0.5979), +0.44 s (+0.66 %) for aPTT (p=0.8704), and +0.72 mg/L FEU (+10.63 %) for D-dimer (p=0.0199). D-dimer showed the highest relative bias but remained below the CD (64.7 %). Degree of haemolysis did not predict the magnitude of interference across the range of haemoglobin (Hb) concentrations studied. CN-3000 flagging grades were constant and correlated free Hb and irrespective of baseline sample PT/aPTT/D-dimer level. Conclusions Mild to moderate spontaneous haemolysis does not produce clinically significant interference in PT, aPTT, or D-dimer. CN-3000 flagging behaviour is predictable and grade-dependent. Re-evaluation of automatic haemolysis rejection thresholds is warranted.
Kaewkhruawan et al. (2026) studied this question.