Combining Lower Extremity Radiographic Markers Begets More Accurate Predictions of Remaining Skeletal Growth. Issue 6 (July 2021)

• Record Type: Journal Article
• Authors: Yu, Kristin E.; Coghill, Griffin A.; Vernik, David; Lapow, Justin M.; Cooperman, Daniel R.
• Is Part Of: Journal of pediatric orthopaedics. Volume 41:Issue 6(2021)
• Publication Date: 2021-07
• Subjects: bone age -- skeletal maturity -- growth predictions -- remaining growth -- leg length discrepancy -- lower extremity; Pediatric orthopedics -- Periodicals; 618.927
• DOI: 10.1097/BPO.0000000000001832
• Languages: English
• ISSNs: 0271-6798

Abstract

Abstract : Background: Radiographic markers of skeletal maturity are vital to the prediction and interpretation of skeletal growth patterns. Accurate predictions of skeletal maturity factor into the management of common musculoskeletal disorders. Bone age is conventionally measured using hand and wrist radiographs. The primary study objective was to optimize skeletal maturity estimates based on the morphology of markers at the hip, knee, and foot rather than conventional upper extremity radiographs. Methods: This was a retrospective analysis of children from the Bolton-Brush collection with anteroposterior radiographs of the hip and anteroposterior and lateral radiographs of the knee and foot, and heights recorded at the time of each radiograph. The percent growth achieved (%GA) was calculated as a function of final patient height. Poor quality radiographs were excluded, leaving 50 patients—32 females and 18 males—and 1068 radiographs for analysis. Skeletal maturity was evaluated using the Oxford bone, O'Connor knee, and calcaneal apophyseal scores. Interrater and intrarater reliability analyses were performed for hip and knee scores. Multiple linear regressions were conducted on these scores and chronologic age as predictors of %GA. Mean differences were calculated between actual and estimated %GAs. All analyses were performed in Prism 8.0. Results: Each lower extremity skeletal maturity score served as statistically significant, independent predictors of %GA, the accuracy and strength of which increased with the addition of chronologic age. The integration of all 3 systems and chronologic age yielded the most predictive, accurate model predictive of %GA, which can be used to determine percent growth remaining. However, this fully integrated system was not statistically superior to the combination of knee and foot scores and knee score and chronologic age, which yield similarly accurate %GA predictions. The hip and knee systems demonstrated good to excellent interrater and intrarater reliability. Conclusions: Integration of bone age scores across different regions of the lower extremity has the potential to facilitate orthopaedic decision-making using radiographs already obtained in the treatment of pediatric musculoskeletal conditions. Level of Evidence: Level IV.