Randomised controlled trial of magnetic-resonance pelvimetry in breech presentation at term.

Background: Pelvimetry is widely used in women with breech presentation at term to select those for whom planned vaginal delivery is appropriate. However, its clinical value has never been established. We evaluated pelvimetry in a randomised controlled trial. The main outcome measures were the elective and emergency caesarean-section rates and the early condition of the neonate.

Methods: Magnetic-resonance (MR) pelvimetry was done on 235 women. The women were then randomly assigned to two groups--for the study group (n = 118), the pelvimetry results were reported to the responsible obstetricians, who used them as the basis for decisions on whether to schedule elective caesarean or trial of labour; for the control group (n = 117), the pelvimetry results were not disclosed until 8 weeks post partum, and decisions about obstetric management were made on the basis of clinical factors only.

Findings: 35 women (15 [13%] study group, 20 [17%] control group) had abnormalities on pelvimetry. The overall caesarean-section rates did not differ significantly between the study and control groups (50 [42%] vs 59 [50%], p = 0.24) but the emergency caesarean-section rate was significantly lower in the study group than in the control group (22 [19%] vs 41 [35%], p = 0.0052). The mean 1 min Apgar scores in the study and control groups were 8.1 and 8.0 (p = 0.93) and the mean 3 min scores 9.5 and 9.4, respectively (p = 0.28). There were no significant differences in the early neonatal outcome for infants born vaginally, by emergency caesarean section, or by elective caesarean section in the two groups, except for a significantly lower Apgar score in the six infants born vaginally to control-group women who had pelvic abnormalities.

Interpretation: The use of MR pelvimetry in breech presentation at term did not significantly reduce the overall caesarean-section rate. However, it allowed better selection of the delivery route, with a significantly lower emergency caesarean-section rate. Neonatal outcome was not compromised by use of the pelvimetry data.