Dongxiang Zhang, Yiling Kang, Lingling Zhu
BACKGROUND Spontaneous preterm birth is a major cause of neonatal morbidity and mortality, yet predicting it in women who are asymptomatic and at low risk is challenging. Cervical length (CL) has limited sensitivity. Shear wave elastography (SWE) quantifies tissue stiffness and may capture premature cervical remodeling. This study aimed to combine SWE parameters with CL to predict spontaneous preterm birth. MATERIAL AND METHODS We retrospectively analyzed 100 asymptomatic women with low risk of preterm birth (singleton, 18-24 weeks) undergoing routine transvaginal ultrasound. The primary outcome was spontaneous preterm birth (<37 weeks). CL and SWE parameters (internal os stiffness [IOS], external os stiffness [EOS], elasticity contrast index [ECI], hardness ratio [HR]) were measured. Multivariable logistic regression and ROC analysis assessed predictive performance, with bootstrap internal validation. RESULTS Twenty-three women (23%) delivered preterm, including 18 spontaneous and 5 iatrogenic cases; sensitivity analysis excluding iatrogenic cases yielded similar results. The preterm group had shorter CL, higher IOS, EOS, and ECI, and lower HR (all P<0.05). Independent predictors were CL (aOR=0.90), IOS (aOR=1.10), and ECI (aOR=1.28). The combined model (CL+IOS+ECI) achieved an AUC of 0.889 (95% CI: 0.791-0.985), sensitivity of 82.6%, and specificity of 87.0%, which was significantly better than that of CL alone (AUC=0.709, P=0.01). Optimism-corrected AUC was 0.881. CONCLUSIONS Combining CL with SWE parameters (IOS and ECI) significantly improved spontaneous preterm birth prediction over CL alone in women who are asymptomatic and at low risk. External validation is needed before clinical use.