PEAK SYSTOLIC OVER PROTODIASTOLIC RATIO AS AN OBJECTIVE SUBSTITUTE FOR THE UTERINE ARTERY NOTCH

Objective To measure the inter-rater agreement for the identification of a uterine artery notch, as well as the association between an obser ved notch and the peak systolic over protodiastolic (A/C) ratio. Desig n Cohort study. Setting Tertiary care university hospital. Population and methods Six hundred and sixty-five nulliparous women in whom 1022 examinations of uterine artery velocity waveforms were performed by pu lsed Doppler at 18 and at 26 weeks of gestation. Agreement between two independent raters was analysed using Cohen's kappa statistics. A/C r atios of flow velocity waveforms with or without a notch were compared . The agreement between A/C values and the presence of a notch was est imated by measuring the surface under the receiver operating character istic (ROC) curve. Results Inter-rater agreement for the identificatio n of a notch was 0.74 (95% CI 0.64-0.83) at 18 weeks and 0.72 (95% CI 0.64-0.80) at 26 weeks. A/C ratios were higher when a notch was presen t (P < 0.0001). The area under the ROC curve was 0.86 (95% CI 0.81-0.9 1) for the placental uterine artery and 0.93 (95% CI 0.90-0.96) for th e nonplacental artery. An A/C value greater than or equal to 2.5 in an y uterine artery had a sensitivity of 88% and a specificity of 86% to detect a notch. Conclusion Although there is no definitive definition of the notch, its detection is reproducible within a center. However, published prevalences between centers in unselected populations vary. The measure of the AVC ratio can serve as an objective substitute.