The Effects of Different Blood Sample Types on Quantitative Detection of Procalcitonin

BACKGROUND: Procalcitonin (PCT) has been recommended and widely used for the prognosis, diagnosis, and monitoring of sepsis. Currently, a majority of PCT detection products are limited to using serum samples, which requires the tedious pre-treatment process, as well as a large sample volume for analysis. Hence, there is an increasing need to replace serum with plasma or whole blood samples. In this work, we evaluated the effects of different blood sample types on PCT quantitative detection by measuring PCT levels in clinically homogenous whole blood, plasma, and serum samples, in hope of extending the application of PCT detection in more sample types. METHODS: Ninety patients from Tong Ren Hospital of Shanghai Jiao Tong University, School of Medicine with different PCT levels volunteered for this study. Clinically homogenous samples, including EDTA- K2 anticoagulant whole blood, centrifuged serum and procoagulant plasma, were collected and analyzed using i-Reader S automatic immunoanalyzer and its supporting reagent kits. Passing and Bablok regression analysis, Bland-Altman plotting, and Kappa test were used for consistency analysis and the determination of consistency intensity, respectively. RESULTS: Correlation analysis showed that PCT concentrations measured from EDTA-K2 anticoagulated plasma samples had a good linear regression relationship with PCT from serum samples, with a slope of 0.8741, r = 0.958, p < 0.05. A similar correlation was observed between whole blood and serum, with a slope of 0.9234, r = 0.965, p < 0.05. A good correlation was found from the quantitative results of different sample types obtained from the same patient. In detail, PCT levels in EDTA-K2 anticoagulant whole blood and plasma are well correlated with that in the serum (r = 0.831, p < 0.05; r = 0.814, p < 0.05). There was no significant difference among different sample types (p > 0.05). Variation in PCT quantification induced by different sample types has no statistical influence on positive/negative decision (p > 0.05). Bland-Altman analysis for assessing PCT concentration agreement showed there was no outlier ratio between whole blood and plasma within the range of the detection system, as well as no outlier between serum and plasma. Kappa coefficient of PCT concentration between serum and homologous EDTA-K2 anticoagulated plasma was 0.8942 (p < 0.001), and for serum and homologous whole blood it was 0.6954 (p < 0.001). CONCLUSIONS: We concluded that quantitative PCT detection can be conducted in different sample types with good correlation and consistency, which means that it is feasible to replace serum samples with whole blood and/or plasma samples for PCT detection in clinical use. These findings reduce the sample volume and turnover time of PCT detection in clinical labs, greatly improving the process of PCT detection and promoting the application of PCT as an important inflammatory biomarker for disease diagnosis.