Clinical Fellow Vanderbilt University Medical Center Nashville, Tennessee, United States
Background: All US children are screened for critical congenital heart disease (CCHD) by pulse oximetry at birth. The only previous national policy analysis of CCHD screening was conducted when just eight states had implemented screening, and it showed a 33.4% decrease in infant deaths from CCHD. No further national studies have been conducted despite widespread adoption. More recent single-center and single-state studies have shown mixed results. Additionally, there is a growing literature that pulse oximetry is less accurate in children with darker skin tones. There has been no published analysis of whether pulse oximetry screening has a differential effect on mortality by race or ethnicity. Objective: We aim to determine whether adoption of CCHD screening policy at the state level reduces mortality from CCHD and whether this effect is modified by race or ethnicity. Design/Methods: This study has been approved by the Vanderbilt University IRB. We will use the period-linked infant birth/death file from the CDC, which contains the month, year, and state of birth for all children as well as maternal/paternal race/ethnicity and cause of death. A data request was approved, and analysis is ongoing. During the study period (2003-2018), there were 64,830,501 births and 8,318 deaths from CCHD. Poisson regression with clustering at the state level will be used to quantify percentage decrease in mortality with policy implementation using difference-in-differences. State and year-month fixed effects will be included. Additional adjustments will be made to account for staggered rollout of state-level policies. Sensitivity analyses will examine outcomes not expected to be affected by screening, including low birthweight deaths. The initial analysis will include all children to quantify the mortality effect of screening policy. Subsequent analyses will be conducted separately by race/ethnicity to determine any effect modification. All data analyses will be completed by January 2024.
This project is supported by grant number T32 HS026122 from the Agency for Healthcare Research and Quality.
