Development and validation of a weight estimation tool for paediatric emergency care.

OBJECTIVE: To develop and validate a weight estimation tool applicable in paediatric emergency care.

METHODS: Using anthropometric data from a computerized database of the primary health care paediatric services, Bilbao (Basque Country, Spain), linear regression models were developed with the objective of estimating weight from height. Subsequently, these models were prospectively validated using a consecutive sample of children attended in the emergency department of two tertiary hospitals. Estimated weights were compared with actual weights, calculating the intraclass correlation coefficient (ICC), mean difference and percentages of estimations falling within 10% and 20% of the actual weight.

RESULTS: Using anthropometric data from 15522 children two weight predictive formulas were developed (Bilbao Formulas). The formulas were validated on a sample of 780 children and estimated weight values with a high degree of intraclass correlation with the real weight (ICC=.93, P<.001) and a mean difference of .63 (SD: 4.3). The percentages of estimations falling within 10% and 20% of the actual weight of the child was 62.2% (95% CI: 58.7-65.6) and 93.1% (95% CI: 91.1-94.8), respectively. Weight estimations were more accurate using Bilbao Formulas than other classical formulas.

CONCLUSIONS: Bilbao Formulas would be a valid tool for estimating weight in children in the emergency department and predict weight more accurately than other more commonly used age-based formulas.