What’s the best way to predict the success of a trial of labor after a previous C-section?

Tessmer-Tuck: The close-to-delivery model without the race variable

Tessmer-Tuck et al developed a model similar to Grobman’s close-to-delivery model, but removed race/ethnicity as a variable and compared it to the accuracy of the Grobman nomogram.⁵ Variables considered in this model were maternal age <30 years (odds ratio [OR]=1.53; 95% CI, 1.00-2.36), body mass index (BMI) <30 kg/m² (OR=1.82; 95% CI, 1.11-2.97), any previous vaginal delivery (OR=3.17; 95% CI, 1.50-6.80), previous vaginal delivery after cesarean (OR=2.24; 95% CI, 1.25-4.18), and absence of a recurrent indication for cesarean delivery (OR=1.81; 95% CI, 1.18-2.76).

The model provided a successful probability of vaginal birth after cesarean ranging from 38% to 98% with AUC of 0.723 (95% CI, 0.680-0.767). When compared with the Grobman model, the AUC for features in the Tessmer-Tuck model was 0.757 (95% CI, 0.713-0.801), similar to the AUC of 0.75 reported in the development of the original model. The predictive accuracy of TOLAC success between 70% and 90% was quite poor at only ±29%.

Grobman’s close-to-delivery scoring nomogram correlates well with actual outcomes and has been externally validated.

Metz: A 5-point scoring tool

Metz et al created a point scoring tool for use on admission to the labor ward, based on 5 variables weighted by degree of correlation with TOLAC success: a history of vaginal birth (OR=2.7; 95% CI, 1.8-4.1), absence of a recurrent indication for initial cesarean delivery (OR=2.0; 95% CI, 1.3-3.1), age <35 years (OR=2.0; 95% CI, 1.1-3.4), BMI <30 kg/m² (OR=1.6; 95% CI, 1.1-2.4), and each point of Bishop score on admission (OR=1.3; 95% CI, 1.2-1.4).⁶

The authors internally validated this scoring tool with an AUC of 0.70 (95% CI, 0.67-0.74), then externally validated the tool with an independent cohort of 585 women and found an AUC of 0.80 (95% CI, 0.76-0.84). In the external validation cohort, TOLAC success rates were 37.4% (95% CI, 27.2-47.5) with a score <10 and 94.4% (95% CI, 90.9-97.8) with a score >16, performing within 8% of the prediction model.

Troyer: A simple 4-point tool

Troyer et al created a simple 4-point scoring tool for use on admission to the labor ward.⁷ The tool’s 4 variables—previous dysfunctional labor, no previous vaginal birth, nonreassuring fetal heart tracing (NRFHT) on admission, and induced labor—were found to reduce the success rate of a trial of labor (P<.05). Dinsmoor et al used this scoring tool in a group of 156 women with an overall TOLAC success rate of 76% (3% higher than Troyer’s group) and found that for labors with a favorable score (0), the tool performed within 5% and for labors with an unfavorable score (≥3), the tool performed within 10%.⁸

Flamm: 5 variables weighted by correlation with TOLAC success

Flamm et al also created a scoring tool for use on admission to the labor ward, based on 5 variables weighted according to degree of correlation with TOLAC success: age <40 years (OR=2.58; 95% CI, 1.55-4.3), history of a vaginal birth (OR=1.53-9.11 depending on where the vaginal birth fell in the woman’s reproductive history), reason other than failure to progress for the first cesarean delivery (OR=1.93; 95% CI, 1.58-2.35), cervical effacement ≥75% on admission (OR=2.72; 95% CI, 2.00-3.71), and cervical dilation ≥4 cm on admission (OR=2.16; 95% CI, 1.66-2.82).⁹ Dinsmoor validated this scoring tool as well in 156 women and found 100% TOLAC success for scores ≥7 (within 5% of the original tool) and 56% TOLAC success for scores ≤4 (compared with 49% for scores 0-2 in the original work).⁸