Exploring Cell Population Data for the Diagnosis and Assessment of Severity in Sepsis: A Preliminary Report.

Abstract

Background: Cell Population Data (CPD), derived from next-generation hematology analyzers, is emerging as a promising tool for diagnosis of early sepsis. This preliminary case-control study assessed the diagnostic utility of CPD and its association with sequential organ failure assessment (SOFA) scores and procalcitonin levels in sepsis patients.

Methods: Seventy-two sepsis patients and 72 age- and sex-matched non-septic controls were enrolled. CPD parameters were measured using a Sysmex XN-1000 analyzer. Univariate and multivariate analyses, including PCA, PLS-DA, and OPLS-DA, were used to distinguish between groups. ROC analysis evaluated diagnostic performance. Spearman's rank correlation assessed associations between CPD, SOFA scores, and procalcitonin.

Results: Several CPD parameters-LY-X, LY-Z, MO-X, MO-Y, NE-WX, NE-WY, NE-WZ, LY-WX, LY-WY, and LY-WZ-were significantly elevated in sepsis. Multivariate analysis identified MO-X, MO-WY, LY-X, and LY-Z as strong discriminators (VIP > 1.15). ROC analysis showed MO-X (> 119.6) had 95.83% sensitivity, 73.61% specificity (AUC 0.876), and IG% (> 0.6) had 83.33% sensitivity, 80.56% specificity (AUC 0.880). IG count (> 40/μL) showed 81.94% sensitivity, 81.32% specificity (AUC 0.859); LY-X (> 79.3) had 93.06% sensitivity, 48.61% specificity (AUC 0.685); and MO-WY (> 752) had 84.72% sensitivity, 45.83% specificity (AUC 0.597). SOFA score correlated with MO-X (r = 0.40, p = 0.007), LY-X (r = 0.40, p = 0.008) and LY-Z (r = 0.39, p = 0.009), while procalcitonin correlated with MO-X (r = 0.30, p = 0.03) and HFLC (r = 0.40, p = 0.005).

Conclusion: CPD is a promising, cost-effective biomarker for diagnosis and severity assessment of sepsis, but inter-equipment variability and current "research use only" status warrant further clinical validation.