Reliable evaluation of a patient's clinical readiness for delivery is critical to improving clinical outcomes during childbirth. Childbirth is a complex mechanical process with many factors, such as cervical length and uterine activity. Incorporating multiple factors into a single mathematical framework may aid the prediction of the timing (e.g. normal vs. protracted) and mode of delivery (e.g. vaginal vs. Cesarean). In this talk, we discuss computational and numerical methods for making better predictions. A stochastic framework is presented in which patient-specific characteristics are described by random variables and childbirth is modeled using stochastic partial differential equations. We then introduce methods for approximating the probability distribution that a patient is in a current (health) state given certain patient-specific characteristics.