A multivariable logistic regression model, with backward elimination, was constructed in order to find independent risk factors associated with BPI and CF. A subgroup analysis involved comparison of features of BPI cases with or without associated shoulder dystocia.