The assessment of developmental status using the Ages and Stages questionnaire-3 in nutritional research in north Indian young children

OBJECTIVE AND BACKGROUND For large epidemiological studies in low and middle-income countries, inexpensive and easily administered developmental assessment tools are called for. This report evaluates the feasibility of the assessment tool Ages and Stages Questionnaire 3.edition (ASQ-3) "home procedure" in a field trial in 422 North Indian young children. METHODS ASQ-3 was translated and adjusted for a North Indian Hindi setting. Three examiners were trained by a clinical psychologist to perform the assessments. During the main study, ten % of the assessments were done by two examiners to estimate inter-observer agreement. During all sessions, the examiners recorded whether the scoring was based on observation of the skill during the session, or on caregiver's report of the child's skill. Intra class correlation coefficient was calculated to estimate the agreement between the raters and between the raters and a gold standard. Pearson product moment correlation coefficient and standardized alphas were calculated to measure internal consistency. PRINCIPAL FINDINGS Inter-observer agreement was strong both during training exercises and during the main study. In the Motor subscales and the Problem Solving subscale most items could be observed during the session. The standardized alphas for the total ASQ-3 scale across all ages were strong, while the alpha values for the different subscales and age levels varied. The correlations between the total score and the subscale scores were consistently strong, while the correlations between subscale scores were moderate. CONCLUSIONS/SIGNIFICANCE We found that the translated and adjusted ASQ-3 "home procedure" was a feasible procedure for the collection of reliable data on the developmental status in infants and young children. Examiners were effectively trained over a short period of time, and the total ASQ scores showed adequate variability. However, further adjustments are needed to obtain satisfying alpha values in all subscales, and to ensure variability in all items when transferred to a North Indian cultural context.