Applying the Small Vulnerable Newborns Classification in the U.S. Context

Introduction

The “Small Vulnerable Newborns” (SVN) framework was introduced in a Lancet series in 2023 to better identify and characterize infants at high risk of morbidity and mortality.^1–4 The key aspect of this framework is to consider gestational age at birth and birthweight under the broader SVN term, rather than treating these characteristics separately.³ This study aims to apply the SVN framework in the United States, by considering the distribution of SVN categories at birth, disparity by race/ethnicity, and the infant mortality rate (IMR) by SVN categories.

Methods

We used the 2021 linked birth-infant death data files provided by the Centers for Disease Control and Prevention’s National Vital Statistics System and included all live births with recorded birthweight and gestational age.⁵ Internal Review Board approval and informed consent were not obtained as the data used was de-identified and made available for public use. The following three mutually exclusive SVN categories were considered:⁴ 1) preterm birth (PTB, born <37 completed weeks) and small for gestational age (SGA; birthweight <10th centile of the international standard of sex-specific birthweight for gestational age);⁶ 2) PTB and non-SGA; and 3) term (born ≥37 completed weeks) and SGA. We also considered a non-SVN category, i.e., term/post-term and non-SGA. We computed the proportion of live births and IMR by SVN category, stratified by race and ethnicity. Finally, we compared the proportion of infants identified as at-risk under the SVN definition that were missed when using the conventional risk definitions separately (PTB, low birthweight [LBW, <2,500 g], and SGA).

Results

A total of 3,666,790 live births in the United States in 2021 were included in the study. SVNs were more likely to be born to women who were older, had lower educational attainment, were Medicaid-insured, had higher body mass index, or were smokers (Table S1). Black women had the highest proportion (23.4%, n=121,412) of SVN births (Figure 1A).

Figure 1.

Distribution of live births and infant mortality rates by SVN categorization, race and ethnicity in the United States, 2021.

A) Live births by SVN categories, race and ethnicity. B) Infant mortality rates by SVN categories, race and ethnicity. C) Comparison of infant mortality rates among infants identified as vulnerable using new (i.e., SVN) vs. old (i.e., SGA, PTB or LBW) definition, by race and ethnicity. LBW, low birthweight; PTB, preterm birth; SGA, small for gestational age; SVN, small vulnerable newborns.

Considering PTB individually would miss 76,061 (4.0%), 45,036 (8.7%), 1,611 (4.5%), 16,587 (7.8%), 43,492 (4.9%) and 6,929 (5.6%) of SVN-births among women identified as White, Black, Indigenous, Asian, Hispanic, and Other, respectively. For LBW, the numbers missed were 127,063 (6.7%), 47,831 (9.2%), 3,125 (8.8%), 17,338 (8.1%), 67,077 (7.5%) and 9,519 (7.8%). Finally, among women identified as White, Black, Indigenous, Asian, Hispanic, and Other, using SGA as an individual marker would miss 161,281 (8.6%), 64,324 (12.4%), 3,992 (11.2%), 16,874 (7.9%), 81,497 (9.2%) and 12,165 (9.9%) of SVN-births, respectively.

Infants born to Black women had the highest IMR (10.5 deaths per 1,000 births) (Figure 1B). While the most severe SVN category (PTB and SGA) comprised only 1.2% (n=44,319) of all births, it constituted between 9.7% (Indigenous, n=26) and 14.4% (Asian, n=113) of the IMR, underscoring that the most severe SVN are at greater risk of infant mortality. The proportion of IMR among non-SVN infants was highest for Indigenous infants (42.4%, n=114) and lowest for Asian infants (20.4%, n=160).

When comparing how well the traditional single-factor vulnerability indicators fared compared with the SVN categories in capturing infant deaths, both PTB and LBW captured almost the same proportion of IMR as the SVN framework (Figure 1C). Using SGA only, however, would miss the majority of IMR identified by SVN.

Discussion

While infants classified as SVN comprised 15.7% of all babies born in the US, 72.6% of infant deaths occurred in this group. More infant deaths were captured using the SVN classification compared with traditional single-factor infant risk-classifications. Thus, the SVN framework may serve as a valuable tool for identifying infants at increased risk of mortality. Black infants were most likely to be SVN and were also at highest risk for infant mortality. We also found that a relatively large proportion of deaths occurred among Indigenous infants who were not classified as SVN. More research is needed to discern the different underlying causes of infant mortality and tailor preventative interventions across different racial or ethnic groups.