High-precision automated bone age: a clinically useful tool in monitoring of treatment effects in children and adolescents.

The clinical value of serial bone age (BA) determinations in children during growth is limited by the manual rater variability (precision 0.63 years). The objective of this work was to determine the precision of automated bone age and bone health index (BHI) measurements by BoneXpert and to establish the time interval at which the automated method can detect a significant treatment effect. The data were from a case-control trial (oxandrolone/placebo) following 90 boys with Klinefelter syndrome (KS) with five visits over 2 years, recording X-rays of both hands. The precision of BA was 0.08 years [0.07; 0.09] 95% CI, leading to a minimal detectable difference of 0.23 years. The effect of androgen treatment on BA was 0.24, 0.77, 1.24 and 1.43 years after 6, 12, 18 and 24 months, respectively. Thus, the effect on BA was detectable by 6 months. In conclusion, automated BA determination is markedly more precise, compared to manual X-ray readings and can detect 0.23 years changes. Automated BA is clinically useful in follow-up in children and adolescents during growth-modulating therapies.