A simplified method for assessing adolescent bone age using a single key bone per age group showed strong agreement with the standard Greulich-Pyle method in a retrospective study. Among experienced raters, approximately 91% of simplified assessments fell within 1 year of Greulich-Pyle estimates, with minimal systematic bias between methods. The findings suggest the approach may serve as a rapid secondary review tool rather than a replacement for standard GP assessment.
Researchers developed the simplified method as a rapid alternative to atlas-based skeletal maturity assessment for large research datasets, resource-limited settings, and secondary clinical review. The study included 117 hand and wrist radiographs from adolescent patients referred for short stature evaluation at Soroka Medical Center in Israel between 2005 and 2015. The cohort included 67 boys and 50 girls who were otherwise healthy and had no skeletal disorders. Radiographs were limited to patients with Greulich-Pyle (GP) bone ages of at least 11 years in boys and 9 years in girls.
The simplified method assigned skeletal age based on a single visually recognizable bone or epiphysis at each maturational stage using reference descriptions derived from the GP atlas. Two experienced pediatric endocrinologists performed duplicate readings using both methods, while two pediatric residents without prior bone age interpretation experience assessed radiographs using only the simplified method. Investigators evaluated agreement between methods, repeatability, and intra-rater and inter-rater reliability.
Among experienced raters, the mean difference between the simplified and GP methods was +0.02 years in boys and −0.14 years in girls. Overall, 92% of male assessments and 91% of female assessments were within ±1 year of GP estimates, meeting the study’s predefined threshold for acceptable agreement.
Repeatability was slightly lower with the simplified method than with GP assessment but remained within what investigators described as clinically acceptable limits for rapid review applications. Intraclass correlation coefficients exceeded 0.9 for most intra-rater analyses across both methods.
Agreement was lower among inexperienced raters. Using the simplified method alone, 77% of female assessments and 85% of male assessments fell within ±1 year of GP estimates, below the predefined threshold for clinical acceptability. Investigators noted that additional training could improve agreement.
The researchers emphasized that the GP method itself demonstrates inherent variability across repeated assessments and stated that the observed performance of the simplified method fell within the expected range of variability for the standard method. They also noted that reliance on a single bone could introduce discrepancies when ossification patterns vary across the hand.
The study was limited by its patient population, which consisted exclusively of adolescents referred for short stature evaluation, potentially limiting generalizability to broader pediatric populations. In addition, inexperienced raters did not perform parallel GP assessments, preventing direct comparison of simplified and standard methods among less experienced readers.
“While not intended to replace detailed GP assessments, the simplified approach offers a practical tool to enhance consistency and efficiency in skeletal age research and secondary evaluations,” wrote lead study author Yehuda Limony. MD, of Soroka University Medical Center and Ben-Gurion University of the Negev in Beer-Sheva, Israel, and colleagues. The researchers also suggested the method may provide a quick check of radiologists' reported GP bone ages in busy clinical practice without requiring a full repeat atlas assessment.
Disclosures: The researchers reported no conflicts of interest.
Source: Clinical Endocrinology
