In this regional population-based study of "healthy" newborns, we found that LGA at birth was associated with hospital re-admission after postpartum discharge for up to 28 days after birth, independent of common risk factors. Hospital admission rate did not differ between SGA and AGA infants.
The association of LGA at birth with neonatal outcomes during the postpartum period has not been thoroughly investigated. LGA has been reported to be associated with various perinatal complications that occur during delivery or in the immediate neonatal period, such as shoulder dystocia, asphyxia, obstetric trauma, metabolic disorders (hypoglycaemia, hypocalcaemia, and severe jaundice), and respiratory disorders (13, 16, 17). However, data regarding hospital re-admission after postpartum discharge is scarce. Only two observational studies have confirmed our findings. In a regional perinatal medical network in the United States, LGA has been found to be associated with a 20% higher odds of re-hospitalization within 2 weeks of delivery (unadjusted OR, 1.20; 95% CI, 1.0, 1.6); the association was no longer significant for re-hospitalization between 3 and 52 weeks of life (25). In a large population-based record linkage study in New South Wales, Australia, term infants born LGA had a higher risk of hospital readmission during the first month of life than AGA infants, with an adjusted RR of 1.10 (95% CI, 1.05, 1.16) (26). However, the underlying mechanisms remain unknown. The fact that LGA infants were more likely to be hospitalized for infection than AGA infants may result from alterations in the immune system and inflammation acquired in utero. Maternal obesity and gestational diabetes, two principal factors of foetal macrosomia, are associated with foetal/placental inflammation and oxidative stress, which, consequently, may affect the structure and function of some systems (27–30). Furthermore, unlike SGA infants, those born LGA are usually considered to be at a low risk of adverse outcomes and, therefore, receive less regular medical follow-up after postpartum discharge. Altogether, these findings suggest that LGA infants are at a high risk of early adverse outcomes and require close medical follow-up after postpartum discharge.
In the present study, we did not find out differences in hospital readmission rate between SGA and AGA infants. This is in agreement with a recent population-based study of term newborns in the United States (25). The absence of risk may be explained by the fact that infants born with severe SGA (i) are usually prematurely extracted and (ii) frequently receive specific medical follow-up. Therefore, infants born at term and SGA are less likely to be affected by severe placental insufficiency with consecutive neonatal morbidities than SGA infants born early. They also usually stay longer in the maternity ward, or in neonatal unit and receive specific medical follow-up during the post-partum period, preventing emergency department visits and hospital admissions. In our study, we excluded newborns with a prolonged initial hospital stay of at least 10 days (i.e. a marker of early severe morbidity) which might restrict the adverse effects of SGA or other neonatal morbidities after hospital discharge. However, this population still needs to be medically followed up because it is at an increased risk of mortality and adverse outcomes persisting in childhood (18–20).
Strengths and limitations
The French National Health Database System does not include some relevant individual factors, such as ethnicity, maternal BMI, breastfeeding, home environment (parental smoking, air pollution), siblings, educational status, and quality of parenting. Some comorbidities and clinical characteristics derived from ICD codes, such as maternal smoking or obesity, could be underreported. Thus, we were unable to identify how these factors and their severity mediate the associations of LGA or SGA on hospital admission after postpartum discharge. However, some of these factors are linked to socioeconomic status which was considered in the analysis. The French National Health Database System accurately registers data on hospitalized newborn and maternal health. All consecutive deliveries in private or public hospitals providing different levels of healthcare are collected in this registered database system, strengthening the findings of the study. We have excluded 12.9% of patients due to coding errors, maternal and neonatal linkage errors, or aberrant data to ensure an optimal quality cohort and to prevent potential selection bias. Finally, this population-based study of a large sample size of "healthy" singleton term newborns provides a relevant representation of a population usually cared for in maternity wards. Nevertheless, our findings need to be confirmed in other societies with different health care systems.