(456) A decision tree model including sonographic measurement of head-perineum distance to predict difficult vacuum delivery

Ultrasound use during birth is a relevant and objective method to help in decision-making in the delivery room. It seems like ultrasound use can predict “difficult birth”. However, there is an absence of consensus concerning the HPD measurement and these studies are essentially single-centre studies.

Thus the objective was to build a decision tree model including sonographic measurement of head-perineum distance to predict difficult vacuum delivery.

Study Design:
This was a secondary analysis of a prospective, multicentric cohort-based study involving 111 private and public maternities in France (Instrumoda study, Gachon et al., AJOG, 2025). Women included were nulliparous women, aged of 18 years or more, with deliveries at 34 weeks of gestation or more, of singleton fetuses in cephalic presentation. In this present study, only women who had a vacuum delivery with measurement of head-perineum distance before operative vaginal delivery, whether the final delivery route, were included. HPD was measured in a frontal transperineal scan as the shortest distance from the outer bony limit of the fetal skull to the perineum. Difficult instrumental delivery was defined as a change of instrument or a c-section.

Results:

2644 cases of vacuum delivery with HPD measurement were included of which 343 (13%) were classified as difficult. HPD was higher in the failure group (43,0 +- 10,1 vs 38,0 +- 11,3 mm, p< 0,001).  Vacuum failures seem to increase with HPD: there is 1,5% of failure in the 0-20 mm HPD group whilst there is 39,1% in the 40-49 mm HPD group. The decision tree was created to establish thresholds according to the fetal position. In the case of an anterior presentation with a DTP< 36.5mm, the risk of vacuum failure was 6%. On the other hand, in the case of a posterior variety with a DTP >30.5mm, the risk of vacuum failure was 22%.

Conclusion:
HPD measurement associated with the fetal position can help predict a difficult vacuum delivery. It will be interesting to evaluate this tree in other populations. (INSTRUMODA Study Group)