The reliability of bone age determination in central European children using the Greulich and Pyle method

This study assessed the reliability of bone age determination in normal cen tral European children using the Greulich and Pyle method and determined th e effects of readers' experience on the measured bone ages. Plain hand radi ographs of 47 children (aged 2 months to 18.8 years) with normal growth wer e analysed by four radiologists (two experienced paediatric radiologists an d two radiology residents). The readers were blinded to the age of the chil dren. The images were re-read by the same readers 2 months later. The mean intraobserver and interobserver variations were lower for experienced reade rs than for radiology residents. However, these differences were not statis tically significant at the 5% level. The difference between the chronologic al age and the measured bone age was -1.5 +/- 7.6 months (p = 0.20) for the experienced readers and -2.7 +/- 10.3 months (p=0.09) for the radiology re sidents. The differences between the measured bone age and chronological ag e were statistically significant (p=0.04) for only one of the two radiology residents. Although the measurements by all four readers underestimated th e chronological age, the differences between chronological age and bone age were within the normal variations of skeletal maturation as reported by Gr eulich and Pyle. Our data suggest that the reliability of bone age measurem ents increases with experience and that the Greulich and Pyle method may be used for central European children.