Accuracy of a smartphone-based 3D imaging tool for height measurement and stunting detection among children 24-59 months of age in Nepal: a validation study.
cross_sectional · Level IV
Where this comes from
- Record sourced from PubMed, PMID 42080151.
- Also identified by DOI 10.1016/j.lansea.2026.100774 and PMC identifier 13129374.
- Licence recorded as CC BY.
- The licence permits redistribution, so the abstract is shown in full and the full text is available from the publisher.
Abstract
Detection of stunting in young children using traditional anthropometry has been a challenge at the community level especially in resource-constrained settings. We evaluated the performance of Child Growth Monitor (CGM), a smartphone-based 3D depth-imaging application, for screening of stunting, and height measurement among children (24-59 months of age) in Nepal. At four health-posts in Rautahat district, Nepal, in October 2024, we conducted cross-sectional validation study comparing CGM's stunting detection against gold standard (manual measurement) and prospective reliability assessment with repeat measurements, four days apart. Eight trained enumerators collected data using calibrated height boards and CGM app. Validity was assessed through sensitivity, specificity and accuracy, reliability through intraclass correlation coefficients (ICC) and Technical Error of Measurement (TEM). Pitman's test evaluated variance. Bias was assessed using Bland Altman statistics. 310 children were recruited, with analyses based on paired subsets (n = 261). Manual measurements identified 36% as stunted, while CGM detected 31·8%. CGM demonstrated 84·0% sensitivity, 98·2% specificity, 96·3% positive and 91·6% negative predictive values, and 93·1% accuracy. For height measurement, the ICC for intra-rater and inter-rater reliability were 0·990 (n = 76) and 0·992 (n = 75) respectively, with intra- and inter-rater TEM of 0·5 cm. Inter-method TEM was 0·7 cm. Pitman's test was significant; Bland-Altman showed mean bias of 0·3 cm. CGM demonstrated accuracy and reliability for field-based stunting detection among children belonging to 24-59 months of age in Nepal. Future research on its use in younger children (6-24 months) and diverse settings will enable its integration into routine growth monitoring. This study was funded by the German Federal Foreign Office as part of the overall funding for the development of Child Growth Monitor app.