Cervix assessment for the management of labor induction: reliability of cervical length and Bishop score determined by residents.

Abstract

To evaluate the reliability of two techniques of cervical ripeness assessment at the beginning of labor induction, as assessed by inexperienced observers. A total of 120 women were prospectively studied at admission for labor induction. Two independent physicians examined consenting women successively but separately. One experienced consultant and one of the six first-year obstetrics residents composed the pairs of observers. The trainees had very limited prior experience with cervical ultrasound scan or evaluating Bishop score. Cervical length and Bishop score were evaluated according to standard methodology. Reliability was described by intraclass correlation coefficients (ICC) and coefficients of variation (COV). Ultrasound cervical length was plotted according to the Bland-Altman method. The median Bishop score and mean cervical length were similar for all observers. Ultrasound cervical length and Bishop score were highly reliable when evaluated by first-year residents. Ultrasound cervical length showed the highest interobserver agreement, with ICC values over 0.9 and COV values of approximately 10% or less for all trainees. The Bishop score was also reliable, but with lower ICC and higher COV values. Bishop score and ultrasound cervical length examination for the management of labor induction are accurate and easy to learn by inexperienced first-year residents. However, cervical length showed higher reliability than the Bishop score.