The predictive value of thromboelastography, routine blood indices, ultrasound parameters, and placental thickness in determining fetal outcome.

Abstract

Objective: To evaluate the predictive value of thromboelastography, routine blood indices, ultrasound measurements, and placental thickness for fetal outcome.

Methods: A retrospective analysis of 218 expectant mothers at our hospital from April 2020 to June 2022 was conducted. Mothers were classified into favorable (n=164) and adverse (n=54) fetal outcome groups. We compared thromboelastography, blood counts, and ultrasound parameters, including placental thickness, between the two groups. Predictive models using lasso regression were developed for individual assessment type and their combinations. Model efficacies were evaluated by ROC curves and Delong's test.

Results: Thromboelastography indicated significantly higher values of R (P=0.004), Angle (P<0.001), and MA (P=0.002) while notably lower K (P<0.001) in the adverse outcome group compared to the favorable outcome group. Peripheral blood analysis showed elevated levels of WBC (P<0.001), CRP (P=0.001), and PLR (P<0.001) in the adverse outcome group. Ultrasound assessments revealed significant increases in S/D (P<0.001), PI (P=0.016), RI (P<0.001), and placental thickness (P<0.001) in the adverse outcome group. The areas under the curve (AUCs) for the thromboelastography (4 features), peripheral blood indices (3 features), ultrasound parameters (4 features), and combined index model (11 features) were 0.774, 0.779, 0.961, and 0.978, respectively. Delong's test indicated that the combined model's AUC did not significantly differ from that of the ultrasound parameters (P>0.05) but was superior to the models based on thromboelastography, peripheral blood indices, and placental thickness alone (P<0.001).

Conclusion: This study underscores the unparalleled predictive value of ultrasound metrics in identifying the risk of adverse pregnancy outcomes, highlighting their critical role in prenatal risk assessment and monitoring frameworks.