Development of prediction models for successful external cephalic version and delivery outcome

Prediction Models for Successful External Cephalic Version: An Updated Systematic Review

“…While the large majority of articles studied maternal weight or BMI, only a few indicated whether this was collected or intended to be a pregravid, peri-ECV, or peripartum BMI. Of the 8 models using some form of maternal weight or BMI, one model used BMI at ECV 46 , one model used the BMI increase from prepregnancy to ECV 39 , two models from the same article used pre-pregnancy BMI 36 , and 4 models were not clear on when BMI was evaluated 30,31,41,44 .…”

“…The remaining 13 new articles used datasets, proposed models, and were performed by researchers that were entirely distinct from the 8 original articles. Of these articles, two were from the same team of authors and used the same dataset; Anand 2019 used the dataset to create a model predicting cephalic position after the ECV, while the Palepu 2021 created a model to predict vaginal delivery 30,31 The remaining 11 articles were completely independent from each other; two of these articles proposed two models each 32,33 , while the remaining 9 proposed a single model each [34][35][36][37][38][39][40][41][42] (Figure 2). Two of the articles had an erratum: one was trivial 36,43 and only modi ed a sentence in the abstract, while the other was a correction of a gure describing the decision tree model 37,44 .…”

“…The results of the study quality assessment based on the signaling questions are shown in Table 4. Out of 17 new articles, 10 studies were identi ed as having low overall risk of bias and high study quality 22,[29][30][31][32]35,37,[40][41][42] , 6 studies were identi ed as having moderate bias and moderate study quality 25,33,34,38,39,43 , and 1 was considered to have high bias and low quality 26 . This is in addition to Velzel et al's assessment of the 8 prior articles of which 4 were low risk of bias and high study quality and 4 were moderate risk of bias and study quality 10 .…”

“…The model types, level development, and model performance are shown in Table 6, 13 models (from 12 articles) of the 25 models were only at the model derivation stage and had no internal or external validation 24,26,28,31,33,34,36,[38][39][40][41]45 . 7 models (from six articles) had internal validation by either random split, bootstrapping, and/or cross-validation, but had no form of external validation 24,29,30,32,37,42 .…”

“…Of the 14 articles that reported neither a calibration plot nor a Hosmer-Lemeshow test, 6 had some other metric that can loosely be used to measure calibration (e.g., likelihood ratios or observed vs predicted chance by score value), while the remaining 8 had none whatsoever. Those articles that reported some form of calibration metric all reported good calibration, except Dong et al which reported good calibration for their ECV success model but not the delivery outcome model 32 . The Newman-Peacock model is the only model that passed external validation in a new location or by another investigator team; neither the original article nor that by the new team reported calibration metrics although the odds of success by score group were reported (Table 7) 25,27 .…”