Clinical and Analytical Performance Evaluation of an Automated Procalcitonin Assay

Abstract Background Procalcitonin (PCT) measurement is useful for guiding antibiotic therapy and risk assessment in lower respiratory infections and/or sepsis. This study evaluated clinical and analytical performance of the Vitros® Immunodiagnostic Products B·R·A·H·M·S PCT assay (Vitros PCT). Methods Precision, limits of blank (LoB), detection (LoD), and quantitation (LoQ) were determined for Vitros PCT, along with method comparison and clinical concordance with the B·R·A·H·M·S PCT™-sensitive KRYPTOR™ assay (KRYPTOR PCT). All-cause 28-day mortality was evaluated according to the change in PCT values (ΔPCT) from day 0 through day 4 in samples from 598 intensive care unit patients with sepsis. Results Comparison of Vitros PCT and KRYPTOR PCT results yielded a Deming regression slope of 1.057, intercept of −0.010, and correlation coefficient (r) of 0.994. Precision analysis demonstrated within-laboratory coefficients of variation for Vitros PCT ranging from 3.1% to 6.4%. The LoD and observed LoQ were determined as 0.007 and 0.013 ng/mL, respectively. Overall agreement between assay methods was 98.5%, 98.0%, 97.4%, and 97.8%, at PCT clinical decision cutoffs of 0.100, 0.250, 0.500, and 2.00 ng/mL, respectively, with Cohen’s Kappa coefficients (κ) > 0.91. ΔPCT values ≤80% vs >80% were associated with increased 28-day-all-cause mortality (P = 0.006). Conclusions Vitros PCT compares well with KRYPTOR PCT, showing excellent agreement at relevant clinical decision cutoffs that have been used for antibiotic decision-making and assessment of risk for sepsis progression. ΔPCT values determined with Vitros PCT were useful for evaluation of 28-day mortality risk in patients with severe sepsis.