Objective: To construct a prediction model for live birth after in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) treatment and single-embryo transfer (SET) after 2 days of embryo culture.
Design: Prospective observational cohort study.
Setting: University-affiliated private infertility center.
Patient(s): SET in 8,451 IVF/ICSI treatments in 5,699 unselected consecutive couples during 1999-2014.
Intervention(s): A total of 100 basal patient characteristics and treatment data were analyzed for associations with live birth after IVF/ICSI (adjusted for repeated treatments) and subsequently combined for prediction model construction.
Main outcome measure(s): Live birth rate (LBR) and performance of live birth prediction model.
Result(s): Embryo score, treatment history, ovarian sensitivity index (OSI; number of oocytes/total dose of FSH administered), female age, infertility cause, endometrial thickness, and female height were all independent predictors of live birth. A prediction model (training data set; n = 5,722) based on these variables showed moderate discrimination, but predicted LBR with high accuracy in subgroups of patients, with LBR estimates ranging from <10% to >40%. Outcomes were similar in an internal validation data set (n = 2,460).
Conclusion(s): Based on 100 variables prospectively recorded during a 15-year period, a model for live birth prediction after strict SET was constructed and showed excellent calibration in internal validation. For the first time, female height qualified as a predictor of live birth after IVF/ICSI.
Keywords: IVF; live birth rate; multiple pregnancy; prediction model; single-embryo transfer.
Copyright © 2016 American Society for Reproductive Medicine. Published by Elsevier Inc. All rights reserved.