ABSTRACT Objectives To compare the analytical and clinical performance of MAGLUMI X3 (Snibe) and Cobas e602 (Roche) assays for sFlt‐1, PlGF, and the sFlt‐1/PlGF ratio in women hospitalized at the Department of Gynecology and Obstetrics, University Hospital in Pilsen, and to evaluate diagnostic accuracy for early‐ and late‐onset preeclampsia (PE). Methods This single‐center real‐world study analyzed 192 serum samples from 164 patients (mean age 31.8 ± 5.0 years). Routine testing was performed on Cobas e602; residual serum was measured on MAGLUMI X3 within 24 h. Precision followed CLSI EP05‐A3/EP15‐A3; agreement used Passing–Bablok and Bland–Altman. Diagnostic performance was evaluated by ROC curves with AUC and 95% confidence intervals (CI); AUCs were compared by the DeLong test. Results Repeatability and reproducibility were acceptable across levels (patient‐sample CVs ~0.7%–2.0% and 1.1%–1.5%; QC CVs ~2.0%–8.9%). Bland–Altman showed a mean bias for sFlt‐1 of −869.9 ng/L (95% CI −1035.6 to −704.2) and for PlGF +6.2 ng/L (−4.6 to 17.0). Spearman's r was 0.99 (sFlt‐1) and 0.97 (PlGF), p < 0.0001. For clinical classification, AUCs were high for both platforms—late‐onset PE: Roche 0.95 (95% CI 0.86–0.99) and Snibe 0.95 (0.87–0.99); early‐onset PE: Roche 0.85 (0.75–0.92) and Snibe 0.84 (0.75–0.91)—with no significant differences by DeLong ( p = 0.71 late; p = 0.32 early). Conclusions MAGLUMI X3 showed acceptable analytical performance and diagnostic discrimination comparable to Cobas e602. Although sFlt‐1 values were systematically higher on MAGLUMI, overall discrimination remained similar across platforms when platform‐specific cut‐offs were applied.
Brož et al. (Sun,) studied this question.