Document Type



The reliability of skeletal maturity assessments of the hand-wrist (Greulich-Pyle) was determined using 60 roentgenograms that were assessed twice by each of two observers. These assessments were made with the roentgenograms masked so that only one bone was visible at a time. The results were compared with those obtained when the same roentgenograms were assessed by the same observers with the whole of each roentgenogram visible. Due to this change in methodology, any influence on the observer of the ages assigned to other bones within a roentgenogram and the appearance of other bones was removed. Thus, truly independent assessments of bone-specific skeletal ages were made. In comparison with the findings when the whole of each roentgenogram was visible, area skeletal ages (means of bone-specific skeletal ages) showed little change in comparability but an increase in replicability. There was a reduction in both the comparability and replicability of bone-specific ages; these were particularly unreliable for the carpals, especially the hamate and triquetral. The ranges of bone-specific ages within groups of bones were increased markedly. These ranges tended to be greater within rays than rows. The present findings suggest that the carpals should be excluded when the hand-wrist is being assessed and that either single ages should be given to groups of bones, e.g., metacarpals, or the bones of the hand-wrist should be assessed in a random order.