# Improving outcomes of periviable births via an enhanced prediction tool

> **NIH NIH R01** · UNIVERSITY OF CALIFORNIA, SAN DIEGO · 2024 · $313,758

## Abstract

PROJECT SUMMARY
 The uncertainty surrounding expected outcomes at periviable gestation leads to several major
challenges. First, clinicians may be unsure of how to counsel families. Second, the lack of clarity makes
families more anxious and causes trauma. Third, it is difficult for both clinicians and families to make the most
informed decisions for the neonate. This is important because making a decision to resuscitate when there are
very poor chances for a good outcome could lead to a futile attempt at resuscitation leading to death, or
potentially a survivor that has severe neurodevelopmental disability. On the other hand, making a misinformed
decision to not resuscitate and proceed to comfort care when there is a good chance of survival without
disability could be even more tragic.
 We will develop and test a modern, comprehensive predictive model for outcomes at periviable
gestation using an existing infrastructure for data collection and implementation, the California Perinatal Quality
Care Collaborative (CPQCC). This population-based network of neonatal intensive care units includes both
academic and community units, which means that results will be generalizable. CPQCC already has an
existing data infrastructure that includes maternal and neonatal data, including follow-up data at 2 years of age,
giving an opportunity to study outcomes that do not exist in similar networks. The setting of the CPQCC allows
for a unique opportunity to both improve on current prediction tools, and to implement and evaluate the
prediction tool in a real-world setting.
 In Aim 1, we will build a predictive model for outcomes in periviable gestation using the most up-to-date
data possible using a broad population-based cohort. This model will be used to build an on-line estimator that
will be used by 20 hospitals across California. In Aim 2, we will evaluate how current practice across ~140
California neonatal intensive care units align with prognostic estimates from the models built in Aim 1. In this
Aim, we will evaluate whether certain patient level factors and hospital level factors appear to fall outside the
norms of typical practice in relationship to prognosis, for therapies provided to the mother prior to birth, and the
infant after birth. In Aim 3, we will implement usage of the estimator across California neonatal intensive care
units in waves of 20 hospitals each over a 1 ½ year period. We will then compare if and how practices change
for periviable gestation infants. In Aim 4, we will conduct a cost-effectiveness analysis of implementing this
estimator in clinical practice. This research will fill several gaps in our knowledge of the use of prediction
models for periviable birth, particularly the gap in our understanding of how using an estimator in practice may
influence and improve clinical decisions and outcomes.

## Key facts

- **NIH application ID:** 10788341
- **Project number:** 5R01HD098287-06
- **Recipient organization:** UNIVERSITY OF CALIFORNIA, SAN DIEGO
- **Principal Investigator:** Henry Chong Lee
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $313,758
- **Award type:** 5
- **Project period:** 2023-06-01 → 2025-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10788341

## Citation

> US National Institutes of Health, RePORTER application 10788341, Improving outcomes of periviable births via an enhanced prediction tool (5R01HD098287-06). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10788341. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*