A Suitable Estimation of Body Surface Area Based on Six Very Used Formulae and on Weight/Height

Giancarlo Ruggieri

This study aims to overcome the difficulties caused by the use of different formulae for estimating body surface area (BSA): misappreciations could happen when comparing data differently indexed, particularly in overweight subjects. A table of heights and weights for males and females, based on the Metropolitan Insurance Tables 1983, was composed so to have by each height a wide range of weights and consistent body mass indices, comprised between 21.7 ± 1.33 and 32.8 ± 0.1 kg/m2 for males and between 20.5 ± 1.32 and 32.8 ± 0.09 kg/m2 for females. Based on the weights-heights tables, six BSAs by six widely used formulae were estimated for each set of height and weight. Their average value represented a reference BSA, which was regressed on the ratio of weight/height, and assumed a measure of a correlated BSA. A very high correlation resulted (R2 > 0.998). Using the equations of the regressions, BSA = x + y * weight/height, the corresponding BSAs were estimated, differing from the reference BSA of –0.0783% for males and of 0.023% for females. A further analysis of this method was processed, using for each height 15 random weights, with an average difference 0.19% ± 0.18 for males and 0.038% ± 0.19 for females. Finally, the differences in percentages between indexations operated using the reference BSAs or the BSAs by equation were evaluated, showing an overall average difference of 0.069% for males and of –0.039% for females. The use of these equations may help the clinician avoid the problems due to the differing indexations because they are based on only a single very suitable BSA and because they include a wide range of weights.