[Inter-observer variability of decision concerning the route of delivery in case of one previous cesarean section and abnormal pelvic measures]

Abstract

Objectives: To study inter-observer variability of decision concerning the route of delivery using pelvimetry in case of one previous cesarean section and abnormal pelvic measures.

Materials and method: Observational study conducted in 2014 in 4 university maternity units among 36 obstetricians. Two groups of obstetricians - as they practiced in a center where pelvimetry was routinely performed (n=12) or not (n=24) - had to choose a route of delivery for 10 clinical cases of women with a single uterine scar and a tight pelvis. The "routine pelvimetry" group had pelvimetry results. The group "no pelvimetry" became aware of pelvimetry results as a second step and had to indicate whether this information changed or not their management. The measurement of the inter-observer variability was estimated by estimating the proportion of agreement according to Grant method.

Results: The proportion of agreements of an attempted vaginal delivery between obstetricians in the group "routine pelvimetry" was 64.7% (95% CI [61-68.5]) and 97.3% (95% CI [96.4 to 98.3]) in the group "no pelvimetry", prior knowledge of pelvimetry results. An attempted vaginal delivery was decided in 77.5% versus 98.7% (P<0.001). After knowledge of pelvimetry results in the group "no pelvimetry" had, the number of attempted vaginal deliver was not different (77.5% vs. 78%, P=0.920).

Conclusion: In women with one previous cesarean section, in case of tight pelvis discovered after pelvimetry, inter-observer variability of decision concerning the route of delivery is increased. Centers that choose to continue using the routine pelvimetry should develop procedures to limit this variability.

Keywords: Delivery; Inter-observer variability; Modalité d’accouchement; Pelvimetry; Pelvimétrie; Previous cesarean section; Utérus cicatriciel; Variabilité inter-observateur.