Abstract
Objective
This study evaluates the analytical validity of capillary blood for Krebs von den Lungen-6 (KL-6) detection using a high-sensitivity chemiluminescent immunoassay (CLIA) on the C5000 analyser and assesses agreement between capillary plasma and venous plasma measurements, with emphasis on methodological feasibility rather than clinical diagnostic performance.
Methods
Analytical performance was assessed in terms of repeatability, accuracy and method comparison. Precision was evaluated using pediatric venous plasma samples at three concentration levels, each tested in 20 replicates. Accuracy was verified using manufacturer standards and external quality controls. Comparative analysis involved 80 paired venous plasma and capillary blood samples, with hematocrit (HCT)-based correction applied to address dilutional effects. Agreement between venous and capillary plasma measurements was evaluated using Passing–Bablok regression, Pearson correlation and Bland–Altman analysis. Pediatric and adult KL-6 levels were also profiled to evaluate age-specific differences.
Results
The C5000 analyzer demonstrated excellent repeatability, with coefficients of variation below 3% across all concentration levels. Accuracy assessment showed minimal bias relative to the standards and controls. Passing–Bablok regression indicated proportional bias before HCT correction, with a slope of 0.769 (95% CI, 0.550–0.945). After correction, the slope was 1.004 (95% CI, 0.882–1.162) and the intercept was −7.03 U/mL (95% CI, −26.97 to 9.76). The mean capillary-minus-venous bias decreased from −48.32 to −3.47 U/mL, and the Pearson correlation coefficient increased from 0.958 to 0.966. Pediatric KL-6 levels were significantly lower than adult levels (
p
Conclusion
High-sensitivity CLIA on the C5000 analyzer supports the analytical feasibility of KL-6 measurement in pediatric capillary blood. HCT correction improved agreement with venous plasma measurements, although further validation is required before the two sample types can be considered interchangeable for individual clinical assessment and before formal reference intervals can be established.