Prediction of preterm birth using the cervical consistency index

Abstract

Objectives: To assess the diagnostic power of a new cervical consistency index (CCI) obtained using transvaginal sonography for the prediction of spontaneous preterm birth (PTB) and to establish reference ranges for this new variable.

Methods: Included in this prospective cross-sectional study were 1115 singleton pregnancies at 5-36 weeks of gestation. Anteroposterior cervical diameter was measured before (AP) and after (AP') application of pressure on the cervix using the transvaginal probe. The index was calculated using the formula: CCI=((AP'/AP) × 100). Cervical length was also measured. The outcomes evaluated were spontaneous PTB before 32, 34 and 37 weeks. Logistic regression and analysis of receiver-operating characteristics (ROC) curves were performed to evaluate the diagnostic power of CCI and cervical length (adjusted for gestational age). Intraclass correlation coefficients (ICCs) and Bland-Altman analysis were used to evaluate intra- and interobserver variability.

Results: In the 1031 women with follow-up, the rate of spontaneous PTB before 32 weeks was 0.87%, before 34 weeks was 2.13% and before 37 weeks was 7.76% (n=80). There were 31 (3.01%) iatrogenic PTBs before 37 weeks. An inverse linear correlation between gestational age and CCI was observed, with regression equation: CCI (in %)=89.8 - 1.35 × (GA in weeks); r(2)=0.66, P<0.001. Cervical length showed an inverse quadratic, though non-significant, relationship with gestational age: CL (in mm)=31.084 - 0.0278× (GA in weeks)(2) + 1.0772× (GA in weeks); r(2)=0.076, P<0.14. The intra- and interobserver ICCs for CCI were 0.99 (95% CI, 0.988-0.994) and 0.98 (95% CI, 0.973-0.987), respectively. The area under the ROC curve for CCI in the prediction of spontaneous PTB before 32 weeks was 0.947, for spontaneous PTB before 34 weeks it was 0.943 and for spontaneous PTB before 37 weeks it was 0.907. For a 5% screen-positive rate, CCI had a sensitivity of 67%, 64% and 45% for prediction of spontaneous PTB before 32, 34 and 37 weeks, respectively, with equivalent values of 11%, 9% and 11% for cervical length.

Conclusions: CCI shows a clear inverse linear relationship with GA. Assessment of CCI is reproducible and is effective in the prediction of spontaneous PTB. This new variable appears to provide better prediction of spontaneous PTB than does cervical length.