Objective: This study was undertaken to use multivariable methods to develop clinical predictive models for the occurrence of uterine rupture by using both antepartum and early intrapartum factors.
Study design: This was a planned secondary analysis from a multicenter case-control study of uterine rupture among women attempting vaginal birth after cesarean (VBAC) delivery. Multivariable methods were used to develop 2 separate clinical predictive indices--one that used only prelabor factors and the other that used both prelabor and early labor factors. These indices were also assessed with the use of Receiver operating characteristic curves.
Results: We identified 134 cases of uterine rupture and 665 noncases. No single individual factor is sufficiently sensitive or specific for clinical prediction of uterine rupture. Likewise, the 2 clinical predictive indices were neither sufficiently sensitive nor specific for clinical use (receiver operating characteristic curve [area under the curve] 0.67 and 0.70, respectively).
Conclusion: Uterine rupture cannot be predicted with either individual or combinations of clinical factors. This has important clinical and medical-legal implications.