Individualised growth charts for preterm infants based on a cohort with healthy neurodevelopment

Abstract

Background & aims: early growth of very preterm infants is associated with later neurodevelopmental outcome. Current growth charts are based on in utero growth rather than a growth pattern associated with good outcomes. This study aimed to generate growth standards using infants who were developing normally.

Methods: data were obtained from the National Neonatal Research Database. Logistic regression identified associations of in-hospital and post-discharge weight gain and head circumference growth with the chance of healthy development at two years. The LMS method was used to construct centile curves reflecting the growth of very preterm infants with a positive developmental outcome. Infants with surgical necrotising enterocolitis or a significant brain injury were excluded from the cohort used to generate growth charts.

Results: growth data were available for 37700 infants, of whom 14120 had a documented developmental assessment. Healthy development was positively associated with three factors: In-hospital weight gain (adjusted OR 1·09 per unit z-score change, 95 % CI: 1·02–1·17), weight gain from discharge to two-year assessment (aOR 1·08, 1·04–1·12) and in-hospital head growth (aOR 1·12, 1.04–1·21). A web app is available (www.bit.ly/preterm-plotter) to generate individualised growth charts for preterm infants, conditioned on their weight and head circumference at birth, to plot their growth and indicate whether their growth was expected to align with that of healthily developing infants.

Conclusion: this study presents a novel method of forming individualised growth charts. It can be implemented using a web app or by integration with clinical information systems to allow an infant's growth to be compared to a cohort of infants with a favourable developmental outcome.

More information

Identifiers

ORCID for James Ashton: orcid.org/0000-0003-0348-8198

ORCID for Mark J. Johnson: orcid.org/0000-0003-1829-9912

Catalogue record

Export record

Altmetrics