BackgroundThe Broselow tape (BT) has been shown to estimate weight poorly primarily because of variations in body habitus. The manufacturers have suggested that a visual assessment of habitus may be used to increase its performance. This study evaluated the ability of habitus-modified models to improve the accuracy thereof.MethodsA post hoc analysis of prospectively collected data from four hospitals in Johannesburg, South Africa, on a population of 1,085 children. Sixteen a priori models generated a modified weight estimation or drug dose based on the BT weight and a gestalt assessment of habitus.ResultsThe habitus-modified method suggested by the manufacturer did not improve the accuracy of the BT. Five dosing and four weight-estimation models were identified that markedly improved dosing and weight estimation accuracy, respectively. The best dosing model improved dosing accuracy (doses within 10% of correct dose) from 52.0 to 69.6% and reduced critical dosing errors from 16.5 to 4.3%. The best weight-estimation model improved accuracy from 59.4 to 81.9% and reduced critical errors from 11.8 to 1.9%.ConclusionThe accuracy of the BT as a drug-dosing and weight-estimation device can be substantially improved by including an appraisal of body habitus in the methodology.