We thank Drs Chubb and Simpson for their interest in our article. They correctly stress the importance of body surface area (BSA) estimation and observer error in the calculation of Z scores.
There is a considerable discrepancy in BSAs calculated using the different published algorithms, including those of Boyd, DuBois and DuBois, Mosteller, and Haycock. This is successfully presented in the graph of the comment of Drs Chubb and Simpson. In our article, we noted that the Haycock algorithm results in the most accurate estimation of BSA in neonates. However, for the calculation of a Z score, BSA should be calculated using the same algorithm applied for the development of the Z -score equation.
Pettersen et al . did not mention the method used for BSA calculation for their normal values, and we did not contact them during the preparation of our report. We assumed they were using the Haycock formula, as it is widely accepted. Chubb and Simpson argued that Pettersen et al . used the formula of DuBois and DuBois. This was indeed confirmed by Pettersen et al . (personal communication, January 2014).
The use of the formula of DuBois and DuBois rather than the Haycock formula results in a predicted aortic valve annular diameter that is 0.11 to 0.32 mm smaller and a predicted mitral valve annular diameter that is 0.17 to 0.49 mm smaller in a neonate of normal length (46–53 cm) ( Figure 1 ). Chubb and Simpson already stated that especially in heavier neonates, this would result in a Z score up to 0.6 lower. This is a significant difference, but probably less than the intraobserver and interobserver variation. Furthermore, the use of confidence intervals for Z scores would be ideal, but these are unfortunately not available yet.
