CPQCC to develop a new prediction model for periviable birth

CPQCC’s Chief Medical Officer, Henry C. Lee, MD, was recently awarded an NIH RO1 grant to develop a data-rich risk prediction model for periviable birth outcomes. The uncertainty surrounding expected outcomes at periviable gestations is one of the biggest challenges for families facing the traumatic circumstance of preterm birth and for the physicians that counsel them. By developing this model, CPQCC aims to improve the care provided to families in these difficult situations.

The project has three aims: to develop contemporary predictive models for periviable birth outcomes and to compare those models to existing prediction tools; to evaluate variation in clinical decision making for periviable infants; and to test whether the prediction provided by a refined outcome estimator influences clinical practice.

Currently, there are several critical gaps in our knowledge that limit more precise counseling of families facing periviable birth. First, while we have refined predictive ability by adding factors beyond gestational age alone, we have not leveraged the full spectrum of maternal, fetal, and hospital data available to optimize risk prediction in a contemporary population. The models that we utilize today often use older data from a select group of hospitals. Moreover, these models are not individualized to the hospital of care. Second, we do not know if using these models impacts the counseling provided by clinicians or if they lead to differences in practice, outcomes, or cost.

As a population-based network that includes both community and academic hospitals and with the infrastructure for statewide data collection and widespread implementation of new practices, CPQCC is uniquely positioned to develop a contemporary predictive model for periviable birth and to close these critical knowledge gaps.