Measurement of the pregnant cervix by transvaginal sonography: an interobserver study and new standards to improve the interobserver variability.

Transvaginal sonography has become an important tool for assessing the gravid cervix uteri, especially in patients at risk for cervical incompetence and preterm delivery. The purpose of our prospective study was to evaluate interobserver variability in measurements of the cervical length and, as a second step, improve the reproducibility and interobserver discrepancy by the introduction of quality control standards when producing and measuring the image of the cervix. Before the introduction of these standards we obtained 46 measurements and observed a mean cervical length (internal to external os) of 33.7 mm (range 26-52 mm) and an interobserver average discrepancy of 3.04 mm (range 0-6 mm). Later, measurements were obtained only when the following conditions were assured and visualized on the screen: (1) the internal os is either flat or is an isosceles triangle; (2) the whole length of the cervical canal can be observed; (3) a symmetric image of the external os can be obtained; and (4) the distance from the surface of the posterior lip to the cervical canal is equal to the distance from the anterior lip to the cervical canal. After the implementation of these quality control standards, 70 measurements were performed obtaining a mean length of 35.3 mm (range 27-51) and an average interobserver discrepancy of 1.24 mm (range 0-4). Thus the measurement error was significantly lower following introduction of quality control standards. Furthermore, the coefficient of variation dropped from 7.1% to 3.3% after the introduction of the guidelines. We believe that these guidelines could be helpful by making the measurements of the cervical length more accurate, reducing the interobserver variability, improving the reproducibility and promoting the role of transvaginal sonography in monitoring the incompetent cervix.