Best Guess method: a further external validation study and comparison with other methods

Validation of the 'Best Guess' weight estimation method on a geographically divergent external sample of children, plus comparison with APLS and Broselow weight estimation methods.

Prospective cross-sectional analytical study at Princess Margaret Hospital Emergency Department. A convenience sample of children aged 0-14 years recruited from May to June 2008. Age, sex, ethnicity, height and actual weight obtained. Agreement between the methods is reported as a comparative mean and distribution of the percentage error, plus the proportion of instances where the error exceeded 20% of the measured weight.

A total of 1235 children were included. The 'Best Guess' method was the most accurate, particularly in children aged 1-4 years (mean percentage error +1.69%). In other age groups it overestimated weight, with mean percentage errors ranging from 3.41% to 6.25%. Across all age groups the Broselow method was most precise, with tendency to underestimate weight across age groups with mean percentage errors ranging from -5.28% to -7.24%. The APLS method was least accurate and precise, with mean percentage errors ranging from -12.61% to -17.36%. Net weight underestimation errors exceeding 20% were associated with increased mean body mass index.

The Best Guess weight estimation method is accurate, especially in children aged 1-4 years. It moderately overestimates weight in other ages. The Broselow method was more precise, whereas the APLS method was the least accurate and precise of all. The ease of use of the Broselow method argues for greater use in the ED and prehospital setting.