Acute Respiratory Failure-Related Mortality Among US Infants, 1999–2023

Technical Summary: Acute Respiratory Failure-Related Mortality Among US Infants, 1999–2023

Problem Statement
Acute respiratory failure (ARF) is a primary driver of mortality in pediatric critical care, with infants under one year accounting for over half of all pediatric ARF cases. Despite the clinical significance of this condition, there is a scarcity of national, long-term epidemiologic data specifically characterizing ARF-related mortality trends and sociodemographic disparities among US infants. Prior research has addressed general pediatric ARF trends or all-cause infant mortality, but lacks granular, 25-year longitudinal data focused exclusively on ARF-associated deaths in the infant population.

Methodology
This study utilized a retrospective, ecological design analyzing data from the US Centers for Disease Control and Prevention (CDC) Wide-ranging Online Data for Epidemiologic Research (WONDER) multiple cause-of-death dataset.

• Cohort: Infants aged <1 year from 1999 to 2023.
• Case Definition: ARF-related deaths were identified using ICD-10 codes J96.0 (Acute respiratory failure), J96.9 (Respiratory failure, unspecified), and P28.5 (Respiratory failure of newborn). The study included cases where ARF was listed as an immediate, underlying, or contributing cause of death.
• Metrics: Crude mortality rates were calculated per 100,000 live births for the overall cohort and subgroups (sex, age, census region). Race and ethnicity-stratified rates were calculated per 100,000 population estimates and were restricted to the period 1999–2020 due to post-2020 revisions in federal racial classification standards.
• Statistical Analysis: Temporal trends were evaluated using Joinpoint regression software (version 5.3.0) to estimate the Average Annual Percent Change (AAPC) and Annual Percent Change (APC) with 95% confidence intervals. Differences between subgroups were assessed using two-tailed two-sample z-tests.

Key Results

• Overall Burden: Over the 25-year period, 67,051 ARF-related infant deaths were identified among 99,053,300 live births, resulting in an overall mortality rate of 67.6 per 100,000 live births.
• Temporal Trends: The overall mortality trend showed a continuous increase, with an AAPC of 0.76% (95% CI: 0.44–1.08; p < 0.001). A distinct upward trend was observed from 2011 to 2023, with an APC of 1.99% (95% CI: 1.50–2.49; p < 0.001).
• Sociodemographic Disparities:
  • Sex: Male infants consistently exhibited higher mortality rates than females (e.g., 84.9 vs. 74.3 per 100,000 in 2023).
  • Age: Neonates (<28 days) had approximately three times the mortality rate of post-neonates (28–364 days).
  • Race/Ethnicity (1999–2020): Black or African American infants had the highest mortality rates (117.7 per 100,000 in 2020), significantly exceeding rates for White, Hispanic/Latino, American Indian/Alaska Native, and Asian/Pacific Islander infants. Notably, Black infants displayed a bimodal trend: a sharp decline from 2006–2012 (APC = -4.25%) followed by a rapid increase from 2012–2018 (APC = 5.35%).
  • Geography: The Southern US had the highest mortality burden in 2023 (86.6 per 100,000), followed by the Midwest, West, and Northeast.

Significance and Claims
The authors claim that this study provides the first national, long-term descriptive analysis of ARF-specific mortality trends in US infants. The primary significance lies in identifying a concerning reversal of historical trends; unlike the previously reported declines in all-cause infant mortality and other pediatric respiratory disease rates, ARF-related mortality has increased continuously since 2011.

The study highlights persistent and prominent sociodemographic inequities, specifically identifying male infants, neonates, Black infants, and those residing in the Southern US as high-risk subgroups. The authors conclude that these findings underscore the urgent need for targeted perinatal and neonatal critical care interventions to address these disparities.

Limitations and Cautions
The paper explicitly maintains a modest stance regarding causal inference. The authors note that as an ecological study lacking individual-level clinical data (e.g., prematurity, birth weight, congenital anomalies), it cannot determine the underlying mechanisms driving the observed trends. Furthermore, the reliance on multiple-cause death certificate codes introduces potential variability in coding practices, and the use of different denominators for race/ethnicity versus other subgroups limits direct comparability. The study does not confirm specific causal hypotheses (such as rural obstetric care contractions or pandemic-related disruptions) but suggests them as areas for future investigation.