ESPE Abstracts (2018) 89 P-P2-240

ESPE2018 Poster Presentations Growth & Syndromes P2 (45 abstracts)

The Validation of an Automated Bone Age Assessment in Girls with Turner Syndrome – A Pilot Study

Ondrej Soucek , Jan Lebl , Klara Maratova , Dana Zemkova & Zdenek Sumnik

Department of Pediatrics, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic

Background: Bone age evaluation is a basic tool to manage the treatment of girls with Turner syndrome (TS). The current standard of care is to involve an experienced medical staff to use the Tanner Whitehouse 3 (TW3) or Greulich-Pyle (GP) method for manual evaluation of the bone age. As this is time consuming and may be partially influenced by the evaluator’s skills, automated systems may prove more efficient.

Objective and hypothesis: The aim of this study was to compare the manual and automated bone age analysis in a pilot group of girls with Turner syndrome of different age. We expected good concordance between the two methods.

Methods: The manual bone age evaluation was performed by an experienced anthropologist while using the TW3 method. The BoneXpert software (Visiana, Denmark) was used for the automated analyses. The difference in the RUS parameter between the two methods was calculated (t-test) and the influence of age and pubertal status was tested (multiple linear regression).

Results: There were 41 girls with TS participating in this study and their mean age was 10.7±3.3 year. The breast stage development according to Tanner’s pubertal scale was 1, 2+3 and 4+5 in 19, 12 and 10 girls, respectively. The mean RUS parameter difference (manual – automated) was +0.6±0.7 year (min. −0.7, max. +2.4) and there was no statistically significant difference between the two methods (P=0.37). In four girls (10%), the automated system computed bone age more than 1.5 years lower than were the manually assessed bone age values. Neither the age nor the pubertal status influenced the difference between the manual and automated bone age measurement.

Conclusions: The automated bone age analysis software produces similar values compared to the manual assessment. Therefore, it keeps promise for more efficiency in daily clinical routine. However, in some girls with TS the extent of underestimation may be of clinical concern. Therefore, validation on larger populations of different diseases is needed to draw the final conclusions and identify the potential pitfalls of the otherwise very convenient endocrinology tool.

Article tools

My recent searches

No recent searches.