A healthy start to life: Using linked administrative data for population-based perinatal and child research

Abstract

Aims: To investigate the role of perinatal factors in determining childhood health, development and school performance. Methods: Probabilistically linked population-based birth, death, hospital, child development and school performance data was used to investigate the association between study exposures and outcomes for a cohort of infants’ liveborn in New South Wales, Australia, 2000-12. Study exposures were gestational age, mode of birth, infant feeding and severe neonatal morbidity (SNM). Study outcomes included children’s 1) health: child hospitalisation for gastroenteritis or any infection; 2) development at ages 4-6 years: special needs; vulnerable or at-risk in one of five developmental domains; or being vulnerable ≥2 domains; and 3) school performance at ages 7-9 years: test exemption; or performing < -1 standard deviation (SD) in reading or numeracy tests. Results: Birth at gestations <39-40 weeks, by labour induction or caesarean section were each associated with an increased risk of hospital admission for gastroenteritis, first and repeat hospitalisation for any infection, being developmentally high risk, or at an increased risk of poorer development in language and cognitive skills, basic literacy and basic numeracy. Formula feeding was associated with an increased risk of hospital admission for gastroenteritis or any infection. Infants with SNM compared to those without, had greater odds of special needs at school age or being exempt from Grade 3 reading and numeracy tests. SNM infants were more likely to have poorer physical and cognitive development and achieve reading and numeracy test scores >1 SD below the mean. Conclusion: We demonstrate that key perinatal factors; gestational age with birth before 39-40 weeks, mode of birth, planned birth, formula feeding and severe neonatal morbidity play an important role in child health and development.