A Digital Health Solution for Child Growth Monitoring at Home: Testing the Accuracy of a Novel "GrowthMonitor" Smartphone Application to Detect Abnormal Height and Body Mass Indices

Thilipan Thaventhiran; Joanna Orr; Joan K Morris; Anne Hsu; Lee Martin; Kate M Davies; Vincent Harding; Paul Chapple; Leo Dunkel; Helen L Storr

doi:10.1016/j.mcpdig.2023.08.001

A Digital Health Solution for Child Growth Monitoring at Home: Testing the Accuracy of a Novel "GrowthMonitor" Smartphone Application to Detect Abnormal Height and Body Mass Indices

Mayo Clin Proc Digit Health. 2023 Dec;1(4):498-509. doi: 10.1016/j.mcpdig.2023.08.001.

Authors

Affiliations

¹ Centre for Endocrinology, William Harvey Research Institute, Queen Mary University of London, United Kingdom.
² Centre for Genomics and Child Health, Blizard Institute, Queen Mary University of London, United Kingdom.
³ Population Health Research Institute, St George's University of London, United Kingdom.
⁴ Department of Paediatric Endocrinology, The Children's Hospital at The Royal London, United Kingdom.
⁵ Department of Health and Social Care, London South Bank University, United Kingdom.
⁶ UCL Digital Experience, Information Services Division, University College London, United Kingdom.

Abstract

Objective: To develop and evaluate a smartphone application that accurately measures height and provides notifications when abnormalities are detected.

Patients and methods: A total of 145 (75 boys) participants with a mean age ± SD of 8.7±4.5 years (range, 1.0-17.0 years), from the Children's Hospital at Barts Health Trust, London, United Kingdom, were enrolled in the study. "GrowthMonitor" (UCL Creatives) iPhone application (GMA) measures height using augmented reality. Using population-based (UK-WHO) references, algorithms calculated height SD score (HSDS), distance from target height (THSDS^DEV), and HSDS change over time (ΔHSDS). Pre-established thresholds discriminated normal/abnormal growth. The GMA and a stadiometer (Harpenden; gold standard) measured standing heights of children at routine clinic visits. A subset of parents used GMA to measure their child's height at home. Outcome targets were 95% of GMA measurements within ±0.5 SDS of the stadiometer and the correct identification of abnormal HSDS, THSDS^DEV, and ΔHSDS.

Results: Bland-Altman plots revealed no appreciable bias in differences between paired study team GMA and stadiometer height measurements, with a mean of the differences of 0.11 cm with 95% limits of agreement of -2.21 to 2.42 cm. There was no evidence of greater bias occurring for either shorter/younger children or taller/older children. The 2 methods of measurements were highly correlated (R=0.999). GrowthMonitor iPhone application measurements performed by parents in clinic and at home were slightly less accurate. The κ coefficient indicated reliable and consistent agreement of flag alerts for HSDS (κ=0.74) and THSDS^DEV (κ=0.88) between 83 paired GMA and stadiometer measurements. GrowthMonitor iPhone application yielded a detection rate of 96% and 97% for HSDS-based and THSDS^DEV-based red flags, respectively. Forty-two (18 boys) participants had GMA calculated ΔHSDS using an additional height measurement 6-16 months later, and no abnormal flag alerts were triggered for ΔHSDS values.

Conclusion: GrowthMonitor iPhone application provides the potential for parents/carers and health care professionals to capture serial height measurements at home and without specialized equipment. Reliable interpretation and flagging of abnormal measurements indicate the potential of this technology to transform childhood growth monitoring.