Validation of prediction model for successful vaginal birth after Cesarean delivery based on sonographic assessment of hysterotomy scar

Objective To validate a prediction model for successful vaginal birth after Cesarean delivery (VBAC) based on sonographic assessment of the hysterotomy scar, in a Swedish population. Methods Data were collected from a prospective cohort study. We recruited non-pregnant women aged 18-35 years who had undergone one previous low-transverse Cesarean delivery at >= 37 gestational weeks and had had no other uterine surgery. Participants who subsequently became pregnant underwent transvaginal ultrasound examination of the Cesarean hysterotomy scar at 11 + 0 to 13 + 6 and at 19 + 0 to 21 + 6 gestational weeks. Thickness of the myometrium at the thinnest part of the scar area was measured. After delivery, information on pregnancy outcome was retrieved from hospital records. Individual probabilities of successful VBAC were calculated using a previously published model. Predicted individual probabilities were divided into deciles. For each decile, observed VBAC rates were calculated. To assess the accuracy of the prediction model, receiver-operating characteristics curves were constructed and the areas under the curves (AUC) were calculated. Results Complete sonographic data were available for 120 women. Eighty (67%) women underwent trial of labor after Cesarean delivery (TOLAC) with VBAC occurring in 70 (88%) cases. The scar was visible in all 80 women at the first-trimester scan and in 54 (68%) women at the second-trimester scan. AUC was 0.44 (95% CI, 0.28-0.60) among all women who underwent TOLAC and 0.51 (95% CI, 0.32-0.71) among those with the scar visible sonographically at both ultrasound examinations. Conclusion The prediction model demonstrated poor accuracy for prediction of successful VBAC in our Swedish population. Copyright (C) 2017 ISUOG. Published by John Wiley & Sons Ltd.