External validation of the Best Guess formulae for paediatric weight estimation

OBJECTIVE: A new method for estimation of weight in children based on their age has been proposed. The present study aims to validate the Best Guess formulae in a new population of children. METHODS: This was a secondary analysis of a database collected for a prospective, observational, cohort study conducted in the Paediatric ED of Sunshine Hospital. Children aged 1-11 years who presented to the ED between 18 August 2005 and 25 February 2006 were included. Actual weight, height, age and ethnicity were obtained. Agreement between estimated weight using the Best Guess formulae and measured weight is reported using mean bias, 95% limits of agreement and proportion within 20% of actual weight. RESULTS: A total of 410 cases were included in the present study. Forty-six per cent were female and median age was 4 years. The mean bias in the 1-5 year group was 0.9 kg, with 95% limits of agreement -3.5 to +5.3 kg. Seventy-six per cent of estimations were within 20% of measured weight. The mean bias in the 5-11 year group was 0.4 kg, with 95% limits of agreement -14.4 to +15.2 kg. In this group, 64% of estimations were within 20% of measured weight. CONCLUSION: The Best Guess formulae performed moderately well in estimating children's weight, but had a tendency to overestimate weight, particularly in children with lower body mass index.