Height is a key component of nutrition assessments in children from limited-resource settings. This study aimed to assess whether handheld digital ultrasound devices for measuring children's height provide comparable accuracy to traditional measurement boards, which are bulky and difficult to transport. We trained 12 health workers to measure the standing height of 222 children aged 2-5 years in rural Lao People's Democratic Republic using both the ultrasound device and measurement board. The Bland-Altman method was used to depict limits of agreement and potential bias. We reported the technical error of measurement (TEM) for precision and accuracy, then assessed these results against the Standardized Monitoring and Assessment for Relief and Transition (SMART) Manual 2.0 and the WHO Multicentre Growth Reference Study (MGRS). The average difference between the ultrasound and board measurements was 0.096 cm (95% limits-of-agreement: 0.041cm, 0.61cm) with a systematic bias of 0.1cm (95% confidence interval: 0.067cm, 0.134cm), suggesting the ultrasound measurements were slightly higher than those from the board. The ultrasound and board TEMs for precision were 0.157cm and 0.091cm respectively. The accuracy TEM was 0.205cm. All TEMs were within SMART and WHO MGRS limits. The ultrasound device is comparable to the measurement board among standing Lao children aged 2-5 years for precision and accuracy TEMs but showed a bias of 0.1cm. Further studies are required to assess whether calibration can minimise this bias and determine the ultrasound's accuracy on recumbent length for infants and younger children.
Read full abstract