Individual Diagnostic Performance of Routine Blood Parameters in Predicting Preterm Birth: A Retrospective Analysis of 2568 Pregnant Women

OpenAlex İngilizce

Preterm birth is a leading cause of neonatal morbidity and mortality. The objective of this study was to test the validity of an inflammatory risk score, originally developed in oncology, for predicting preterm birth, and to determine the individual diagnostic performance and risk association of routine antenatal blood parameters in predicting this outcome. This retrospective case-control study was conducted with data from pregnant women who delivered at Giresun Gynecology and Pediatrics Training and Research Hospital between January 1, 2019, and December 31, 2024. The study included 1270 pregnant women with singleton pregnancies who delivered preterm (case group) and 1298 randomly selected women who delivered at term (control group). Demographic and laboratory data of the groups were compared. The diagnostic performance of the external oncology score and each laboratory parameter (e.g., CRP, WBC, NLR, BUN, RDW) was evaluated using ROC curve analysis (AUC). The association of each parameter with preterm birth was examined using univariate and multivariable logistic regression (Adjusted Odds Ratio - aOR) adjusted for maternal age. To evaluate the cumulative predictive power, a combined risk model was constructed. No demographic differences were observed between groups ( p > 0.05). All laboratory parameters showed significant differences ( p < 0.001). In multivariable models adjusted for maternal age, WBC (aOR = 1.963), BUN (aOR = 1.321), and CRP (aOR = 1.173) were independently associated with preterm birth. The combined risk model achieved an AUC of 0.874, significantly outperforming individual markers. Optimal cut-off values were > 9.36 for WBC and > 20.00 for BUN. External models, such as oncological risk scores, may exhibit limited performance when directly applied to different physiological states like pregnancy. In contrast, simple and routine blood tests such as WBC, BUN, and CRP demonstrate strong independent associations with preterm birth, even after adjusting for maternal age. The high diagnostic performance of the combined model suggests that these parameters can be integrated into clinical screening protocols for early risk assessment.