Abstract
PURPOSE: One method of predicting leg-length discrepancy at maturity is the Moseley straight-line graph. Beumer et al developed an alternative graph, using a more modern Dutch population. The purpose of this study was to compare the prediction accuracy of these two graphs in a cohort of patients treated at our institution using epiphysiodesis. METHODS: We identified 76 patients treated using epiphysiodesis for leg-length discrepancy who were followed to maturity and had adequate preoperative radiographic assessment for straight-line graph construction. We compared predicted long leg length (after epiphysiodesis), short leg length, and residual leg-length discrepancy to actual outcome for both methods, using both chronological and skeletal ages. RESULTS: Both methods were more accurate using skeletal age rather than chronological age. The Rotterdam graph showed modest improved accuracy compared to the Moseley graph in developmental aetiologies and in Hispanic patients. Using a difference of one centimetre in prediction error as clinically relevant (long leg [after epiphysiodesis], short leg, and leg-length discrepancy in each of the 76 patients, 228 predictions), we found comparable predictions in 171, more accurate prediction using the Rotterdam in 32, and using the Moseley in 25 predictions. CONCLUSIONS: Straight-line graphs provide a generally more accurate prediction of leg lengths at maturity by virtue of multiple preoperative evaluations. The Rotterdam straight-line graph was equal to or superior to the Moseley graph in most patients in this cohort. Use of skeletal age resulted in more accurate predictions than chronological age. Clinicians should remain familiar with the concept and use of the straight-line graph. LEVEL OF EVIDENCE: III, case-control study.