Feasibility of caregiver-administered anthropometric measurements of children under age 5: evidence from Zambia

Abstract

Abstract Background Accurate measurement of children’s anthropometry is of central importance for the assessment of nutritional status as well as for the evaluation of nutrition-specific interventions. Social distancing requirements during the recent Covid-19 pandemic made administration of standard assessor-led measurement protocols infeasible in many settings, creating demand for alternative assessment modalities. Objective To assess the feasibility and reliability of caregiver-administered anthropometric assessments of children under age 5. Design We compared standard and caregiver-administered assessments within an ongoing nutrition trial in Zambia (NCT05120427). We developed a “no-contact” protocol whereby trained staff verbally instruct caregivers from an appropriate distance to measure the height, weight and MUAC of their children. We captured measurements of height, weight and MUAC among a sample of caregivers and infants in Zambia using both the “no-contact” protocol and a standard assessor-led protocol. We analyzed each anthropometric variable, comparing means between protocol group, the proportions yielding standardized z-scores outside the plausible ± 6SD range and the proportions of children classified stunted, underweight and wasted. Results Anthropometric measurements were captured for 76 children using both the no-contact protocol and the standard protocol. An additional 1430 children were assessed by the standard protocol only and an additional 748 children by the no-contact protocol only. For the 76 children measured by both methods, we find no differences in average height, weight and MUAC between caregivers and interviewer assessments. The estimated kappa for the binary stunting and underweight classifications were 0.84 and 0.93, respectively. In the larger samples measured only following one protocol, we find no differences in average outcomes after adjusting for child, caregiver and household characteristics. Conclusions Anthropometric measurement protocols administered by caregivers with verbal instruction from trained assessors are a promising alternative to standard protocols in situations where study staff are unable to come in close contact with study participants. Clinical trials registration This study was conducted within a larger trial registered at clinicaltrials.gov as trial NCT05120427. https://clinicaltrials.gov/ct2/show/NCT05120427.