RESUMEN
STUDY DESIGN: Cross-sectional retrospective analysis. OBJECTIVE: To examine the degree of correlation between thoracic dimension outcome measures and pulmonary function in early-onset scoliosis (EOS). SUMMARY OF BACKGROUND DATA: Change in thoracic dimension (TD) measurements and spine length are commonly reported outcome measures after treatment for EOS. Although ultimately improving or maintaining pulmonary function is the goal of EOS treatment strategies, it is unclear whether commonly reported 2-dimensional TD measurements represent good predictors of pulmonary function. METHODS: A cross-sectional analysis of patients including all diagnoses obtained from 2 EOS databases containing TD measurements and pulmonary function data was performed. Relationships between individual TD measurements and pulmonary function measurements were assessed using the Pearson correlation analysis. TD measurements (pelvic inlet width, T1-T12 height, T1-S1 height, and coronal chest width) and standard pulmonary function measurements were compared. TD percentiles normalized for pelvic inlet width were also calculated and correlated with pulmonary function measurement percentiles. Univariate and multivariate linear regression analyses determined whether TD measurements could predict pulmonary function. RESULTS: There were 121 patients (65 females, 56 males) in the study. Mean age at evaluation was 9.3 years (range, 2.7-18.1 yr). T1-T12 height, T1-S1 height, maximal chest width, and pelvic inlet width were all significantly correlated with forced air volume expelled in 1 second, total forced air volume, and total lung capacity (correlation coefficients [r] 0.33-0.61; all P<0.001). T1-T12 predicted percentile (normalized for pelvic width) was significantly correlated with forced air volume expelled in 1 second and total forced air volume predicted percentiles (r=0.32, P<0.001 and r=0.27, P=0.004, respectively). Regression analysis determined that T1-T12 percentile was a significant predictor of forced air volume expelled in 1 second percentile and total forced air volume percentiles. Regression analysis found no predictive factors of total lung capacity percentile. CONCLUSION: Traditional 2-dimensional TD measurements (T1-T12 height) used to measure outcomes in EOS can be used as weak predictors of pulmonary function outcome. However, better outcome measures need to be developed, such as 3-dimensional and dynamic measurements. LEVEL OF EVIDENCE: 3.