Intrapartum Sonography for Difficult Labor.

Abstract

Difficult labor - protracted or arrested dilatation in the first stage, arrest of descent in the second stage, and the decision whether to attempt an assisted vaginal birth - accounts for a large proportion of intrapartum cesarean deliveries and for most of the maternal and neonatal morbidity attributable to instrumental birth. The decisions made in these situations rest almost entirely on three pieces of information obtained by digital vaginal examination: the position of the fetal occiput, the attitude of the fetal head, and the station of the leading bony point. Digital examination is uncomfortable, must be repeated, and is least reliable precisely when the clinical stakes are highest. Caput succedaneum, molding, deflexion of the head, and a high presenting part obscure the sutures and fontanelles on which the examination depends, and the ischial spines are palpated rather than seen. Studies comparing digital examination with sonography before assisted vaginal birth have reported discordant assessment of occiput position in approximately one fifth to one third of examinations, and digital estimation of station is similarly imprecise, particularly in the presence of substantial caput. Sonography in labor is performed with the curvilinear transducer already available on most labor wards, requires a small number of images, and is better tolerated by women than repeated digital examination. This article is written for general obstetricians, laborists, and residents who manage labor but who do not routinely scan during labor, rather than for clinicians already practicing intrapartum sonography. It is organized around three clinical situations in which a scan may change management - an uncertain fetal head position, slow or arrested labor, and the period immediately before an assisted vaginal birth - and around five clinical questions that arise within them. The five questions, and the sonographic answer to each, are: what is the position of the fetal head, answered by transabdominal identification of the midline cerebral echo, orbits, cerebellum and cervical spine; if the occiput is posterior, is spontaneous rotation likely, suggested by the orientation of the fetal spine; is the head well flexed and is it asynclitic, quantified by the occiput-spine angle in occiput-anterior and occiput-transverse positions and by the chin-to-chest angle in occiput-posterior positions; how low is the head and has it engaged, quantified transperineally by the angle of progression and the head-perineum distance; and is the head descending during pushing, assessed by serial measurement and by the direction of the fetal head. The evidence supporting these measurements is uneven and should be described accurately. Sonographic determination of occiput position is more accurate than digital examination, two randomized trials show that adding a scan improves the accuracy with which the position is known, and one randomized trial reported more accurate placement of the vacuum cup. The measurements of station and attitude are reproducible and are consistently associated with the mode of delivery, but their published thresholds derive from modest cohorts, are correlated against digital station, and are influenced by transducer pressure, caput, molding, parity and occiput position. No trial has yet shown that acting on these measurements reduces maternal or neonatal morbidity. Sonography should therefore be used as an adjunct when clinical examination is uncertain or when a more accurate assessment could alter management, and not as a replacement for it.