Assessing newborn scoring with each resuscitation (ANSWER): Protocol for identifying and testing an Apgar score for the 21st century

Abstract

Background: The status of infants in the minutes after birth has been summarized by the Apgar score for the last 70 years and is applied at most medically attended deliveries around the world. It has not, however, adapted to changes in neonatal resuscitation over the decades. There are issues with application to premature newborns, how to account for the interventions outlined in the Newborn Resuscitation Program (NRP), inter-rater reliability, and local custom. Developing a modern newborn assessment score will require a series of steps, the first of which is to identify which observations or data can best differentiate a normal newborn transition from an abnormal one.

Methods: Video recordings of at least 35 random normal (meeting NRP goals for heart rate and saturation without intervention, normal physical exam and requiring only normal postnatal care) and 35 random abnormal (not meeting at least 2 normal criteria) from 7 centers participating in the VERIFI study. The presence or absence of observable elements, including vital signs, appearance, and interventions will be recorded every 30 seconds for the first 5 minutes of life. Observations, as well as any changes over time or after intervention will be compared between normal and abnormal, and those that are significant and independent based on logistic regression will become candidate newborn assessment score components.

Expected results: There are likely to be ten or more observations/elements from videos of the first five minutes of life that will differ between normal and abnormal newborns, and these will be tested in all combinations to identify the 1-3 score sets that will be applied to a new set of VERIFI videos to identify which has the best sensitivity/specificity.

Conclusions: Bringing newborn assessment into modern practice will build on the legacy of Dr. Apgar. A series of steps, beginning with identifying the observations/elements that best identify newborns who may need further care can lead to a universal validated tool for the 21st century.