Abstract:
|
Measuring cervical dilation in the late stage of pregnancy is a commonly used technique for monitoring the progression of labor. Recent statistical methodology has been developed to address the analytical challenges for such data when only a single labor curve is observed on each woman (McLain and Albert, 2014, Biometrics). These challenges include conducting valid inference and prediction when there is not a time zero (i.e., when women enter the hospital at different stages of their labor). Motivated by the NICHD Consecutive Pregnancy Study (CPS), a unique cohort study that collected repeat labor data on over 50,000 women, we propose new methodology for analyzing labor curves across multiple pregnancies. We propose a latent modeling approach that characterizes serial correlation in the labor process across multiple pregnancies using an Ornstein-Uhlenbeck process. We employ Bayesian methodology (MCMC) for parameter estimation and prediction. The methodology was used in analyzing the CPS data, and in developing a predictor for labor progression that can be used in clinical practice.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.