Addition of Midthigh Circumference Improves Predictive Ability of Broselow Tape Weight Estimation.

Abstract

This study aimed to improve the accuracy of Broselow tape (BT) weight prediction by adding midthigh circumference (MTC) and to compare and standardize the methods for measuring MTC. This prospective, observational study was conducted in a pediatric emergency department of a tertiary care children's hospital. Children up to 12 years of age presenting to emergency department were included. Children were excluded if obtaining the data would interfere with their acute management. The data collected included MTC, height, BT weight, and actual weight. Three models were built. Broselow tape-only model and MTC plus BT model used 2 methods for MTC measurements: visual approximation of thigh midpoint (visual MTC1) and Centers for Disease Control and Prevention (CDC)-defined method (CDC MTC2). A total of 430 children were enrolled during pre-coronavirus disease era. Of these, 234 were boys (54.4%) and 196 (45.6%) were girls. Data were categorized into children younger than 2 years (224 [52.1%]) and older than 2 years (206 [47.9%]). African American constituted 250 (58%); White, 136 (31%); and unspecified, 27 (6%). For the entire cohort, both models that included MTC were significantly better at weight prediction with larger adjusted R2 (visual MTC1, 0.921; CDC MTC2, 0.928) and smaller root mean squared (RMSE) (visual MTC1, 2.70; CDC MTC2, 2.56) compared with BT-only model (adjusted R2 = 0.843; RMSE, 3.80). Midthigh circumference plus BT models performed even better in children older than 2 years compared with BT-only model with adjusted R2 (visual MTC1, 0.859; CDC MTC2, 0.872 vs 0.616) and RMSE (visual MTC1, 3.18; CDC MTC2, 3.03 vs 5.27). The inclusion of MTC with BT resulted in a more accurate weight prediction in children especially greater than 2 years old. Midthigh circumference model using CDC-defined method was slightly better predictor of actual weight than visual approximation.