From kit inserts to daily practice: Six sigma and maximum allowable uncertainty in coagulation analysis

Accurate laboratory test results are essential for patient care. While manufacturers provide validated performance metrics such as the coefficient of variation and bias in kit inserts, maintaining such precision in daily laboratory use is often difficult. This study compared the analytical performance of a Sysmex CS-2500 coagulation analyzer in routine use with manufacturer-declared specifications using the Six Sigma and maximum allowable uncertainty methods. Data from June to November 2024 were collected at Ümraniye Training and Research Hospital. Coagulation tests (prothrombin time [PT], International normalized ratio [INR], and aPTT) were assessed using internal quality control and an external quality assessment scheme. Sigma metrics were calculated per CLIA 2024 and maximum allowable uncertainty values were derived using biological variation data. Laboratory results were compared with kit insert values and evaluated according to analytical performance specifications. SigmaLAB scores were ≥ 5 in all months except July due to external quality assessment-related bias, while SigmaMD remained > 6. In Mu analysis, PT (June-August) and INR (August) fell below minimum analytical performance specifications limits in MuLAB, whereas aPTT consistently met desirable limits. MuMD showed PT within minimum, INR borderline minimum, and aPTT borderline optimal limits. Six Sigma provides retrospective insights, while Measurement Uncertainty (Mu) allows continuous performance monitoring. Combining both offers a full picture of analyzer performance. Together, they provide a more complete assessment, revealing that despite excellent manufacturer data, field results showed bias-related limitations.