6.6 Vaginal Birth after Caesarean Section (VBAC) Predictor Models: which is the best model and can it save money?

    loading  Checking for direct PDF access through Ovid



To assess the performance of three different statistical models in predicting successful VBAC. The statistically most reliable model was subsequently subjected to validation testing in a local antenatal population and a cost saving analysis was performed.

Materials and methods

The study population included all women who had attempted trial of labour after previous Caesarean Section (TOLAC) within a regional UK obstetric unit, between April 2010 and April 2012. A retrospective observational study was performed with study data collected from the Northern Ireland Maternity Service Database (NIMATs) and subjects’ personal medical chart. Statistical analysis was performed using Microsoft EXCELTM 2007 (v.12)and IBM SPSS statisticsTM v. 20.0 (2011).


Our study included 385 TOLAC subjects. Of the three predictor models evaluated, area under the curve (AUC) calculations determined the Grobman model, was statistically the most reliable (AUC = 0.724 p < 0.05). The correlation between observed and predicted outcome in our study cohort using the Grobman model was high (R2 = 0.88), validating the model within our population. Had this model been utilised ante-natally, it may have led to a 3.4% reduction in costs over two years had women with a 72% predicted probability of successful VBAC been selected.


The Grobman model could potentially be utilised within the UK antenatal population to give women and clinicians an informed choice when deciding on mode of delivery after 1 previous CS. This could lead to potential cost savings.

Related Topics

    loading  Loading Related Articles