Endotracheal tube selection in children: A comparison of four methods

Abstract

To determine the accuracy of four methods of endotracheal tube size selection in the pediatric population. Prospective, blinded comparison. The Children's Hospital of Philadelphia. Two hundred thirty-seven children aged 1 month to 9 years old undergoing elective surgery requiring endotracheal intubation. Consecutive sample. Four methods of determining proper endotracheal tube size in children were compared. These methods included direct comparison with the width of the fifth finger, direct comparison with the diameter of the fifth finger using a ring-sizing device, direct comparison with the width of the fifth fingernail, and estimation using a formula ([age in years + 16]/4). In infants, a 3.0-mm (internal diameter) endotracheal tube was predicted for those 3 months of age and younger, and a 3.5-mm endotracheal tube was predicted for those from 3 to 9 months of age. The size of the endotracheal tube used in the operating room was recorded, as was the "air leak" around the tube. An appropriately sized endotracheal tube was determined by an air leak with ventilation pressures between 5 and 40 cm of water. Direct comparison using the width and the diameter of the fifth finger predicted an endotracheal tube between 1 mm smaller and 0.5 mm larger than that used by the anesthesiologists in 11% and 14% of patients, respectively. The age-based formula predicted an endotracheal tube size in this range in 97.5% of patients, and direct comparison with the width of the fifth fingernail predicted an endotracheal tube in this same range in 91% of patients. These findings were consistent within all age groups studied. Neither fifth finger width nor fifth finger diameter accurately predicts proper endotracheal tube size in most children. A more accurate estimation can be made using the age-based formula, but when the child's age is unknown or when calculation is awkward or impossible, an accurate estimate can be made using the width of the fifth fingernail.