Background: Aortic sized index (ASI) defined as aortic root size/body surface area (BSA) is used to provide information on dissection risk in Turner Syndrome (TS). There are multiple equations for estimation of BSA. The impact of using a different BSA equation for calculation of ASI is unknown.
Method: We calculated BSA of 114 TS girls from 2273 outpatient visits using Dubois, Mostellar, Haycock, Gehan, Boyd and Furqan formulae. BSA estimation with Dubois [original equation] was used as gold standard. We compared ASI using the different BSA equations in 130 other girls, median age 16.9 years (1.1, 58.2), with aortic root measurements on echocardiogram.
Results: All BSA formulae were highly agreeable, with Mosteller (mean error −0.007, −0.021 to 0.007), and Haycock (mean error 0.001, −0.014 to 0.016) having all estimations accurate to within 5%. However, ASI calculated using all BSA equations underestimate ASI compared to when BSA was estimated with Dubois, with mean error ranging from −0.033 [Mostellar] to −0.091 [Furqan]. Up to 2% of girls in the high risk ASI [using Dubois BSA] will be reclassified as moderate risk [using Boyd and Gehan BSA]. Up to 8% of girls in the moderate risk ASI [using Dubois BSA] will be reclassified as low risk [using Boyd and Furqan BSA]. In multiple linear regression of factors affecting ASI error, height and weight were significant independent factors (P< 0.0001).
Conclusion: Whilst the limits of agreement between the five equations for estimation of BSA compared with Dubois is high, our study demonstrated for the first time that aortic dissection risk using ASI may be underestimated in some TS girls simply by using the other BSA equations. Given the significant clinical implications, we believe that more accurate and robust methods of evaluating dissection risk in TS are needed.
10 - 12 Sep 2016
European Society for Paediatric Endocrinology