Incremental prognostic value of the fibrinogen-to-albumin ratio for adverse perinatal outcomes in preeclampsia: a dual-center retrospective cohort study.

OBJECTIVE

This study aims to evaluate whether the fibrinogen-to-albumin ratio (FAR) provides incremental prognostic value for identifying preeclamptic patients at high risk of adverse peripartum events, defined as a composite adverse perinatal outcome (CAPO) that serves as a surrogate for severe disease burden, and to explore its potential utility in guiding anesthesia-relevant decisions.

METHODS

This dual-center retrospective cohort study included 924 preeclamptic patients from two tertiary referral hospitals. Patients from center A (n = 680) were randomly allocated in a 7:3 ratio to a training cohort (n = 476) and an internal validation cohort (n = 204). Patients from center B (n = 244) constituted an independent external validation cohort. CAPO was defined as the occurrence of placental abruption, preterm birth, fetal growth restriction, fetal distress, neonatal respiratory distress syndrome, 5-min Apgar score below 7, neonatal intensive care unit admission, or perinatal death. Candidate predictors were identified using forward stepwise logistic regression. Model A (the base model) incorporated established predictors including the sFlt-1/PlGF ratio. Model B (the extended model) was developed by adding FAR. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), calibration plots, and Brier scores. Incremental value was quantified by the change in AUC (ΔAUC), continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI).

RESULTS

In the training cohort, model A achieved an AUC of 0.852 (95% confidence interval [CI] 0.818-0.887), whereas model B demonstrated significantly improved discrimination with an AUC of 0.888 (95% CI 0.859-0.917; P < 0.001). Model B also showed favorable calibration (Hosmer-Lemeshow, P = 0.594) and a lower Brier score (0.134 vs. 0.153). At the optimal cutoff value of 0.135 for FAR (determined by the Youden index), model B achieved a sensitivity of 0.864 and a specificity of 0.740 for predicting CAPO. In the external validation cohort, model B maintained robust discrimination (AUC 0.856, 95% CI 0.808-0.904) and outperformed model A (AUC 0.798, 95% CI 0.740-0.857; P = 0.003). Across the entire cohort, the addition of FAR resulted in a ΔAUC of 0.010 (95% CI 0.002-0.019; P = 0.013), a continuous NRI of 0.397 (95% CI 0.268-0.526; P < 0.001), and an IDI of 0.022 (95% CI 0.012-0.032; P = 0.004). Subgroup and sensitivity analyses confirmed the robustness of these findings.

CONCLUSIONS

FAR provides statistically significant incremental prognostic value for predicting CAPO, a composite of severe peripartum events. From an anesthesiology perspective, FAR ≥0.135 alerts to coagulopathy and systemic inflammation, supporting cautious neuraxial anesthesia, early invasive monitoring, and escalated perioperative care.