Presented at the Neonatal Society 2018 Summer Meeting.
Andrews ET1, Ashton JJ2,3, Pearson F1, Beattie RM2,4, Johnson MJ1,4
1 Department of Neonatal Medicine, Princess Anne Hospital, University Hospital Southampton NHS Foundation Trust, UK
2 Department of Paediatric Gastroenterology, Southampton Children’s Hospital, UK
3 Human Genetics and Genomic Medicine, University of Southampton, UK
4 4National Institute for Health Research, Southampton Biomedical Research Centre, University Hospital Southampton NHS Foundation Trust and University of Southampton, Southampton, UK
Background: Measurement of length and head circumference (HC) in addition to weight is vital in assessing the nutritional status of preterm infants. Current anthropometry represents an interruption to preterm infants and may not be possible in unstable infants. Hand-held 3D scanning has the potential to perform bedside anthropometry (length and HC) in a less invasive manner. We aimed to evaluate the feasibility and performance of 3D scanning as a ‘non-touch’ measuring technique for routine anthropometry.
Methods: Preterm infants born before 30 weeks gestation were recruited from a single neonatal unit. HC and length were measured both manually and by a handheld 3D scanner at recruitment and weekly until discharge. The two methods were compared using the Bland-Altman method and linear regression. This study received ethics approval from an NHS Research Ethics Committee (Oxford A, ref 14/SC/1275).
Results: 17 infants had scan images taken over a 4-month period (87 separate length and 67 HC scan measures with manually taken reference measures). The mean (95%CI) difference between manual and scanner HC measurements was 0.18cm (-0.06 to 0.42cm), the mean percentage difference with all values expressed as positive was 3.16% (2.33% to 4.00%). The mean difference for length measures was 0.27cm (0.03 to 0.54cm) the mean for length percentage difference with all values expressed as positive was 3.24% (2.69 to 3.79%). Bland-Altman plots showed reasonable agreement between the two methods. Linear regression demonstrated a high correlation between scanner and manual measurements of HC (r=0.96, Figure 1a) and length (r= 0.96, Figure 1b).
Conclusion: These data show a high correlation between measurements gathered from 3D scan images and standard anthropometry. This suggests that 3D scanning could represent a feasible, accurate and practical way of monitoring the growth of preterm infants with minimal handling and without interruption to developmental care.
Corresponding author: edwardtandrews@gmail.com