O1051: Influence of Pelvic Height Difference in Patients with Adolescent Idiopathic Scoliosis Undergoing Spinal Fusion

Background: Patients with adolescent idiopathic scoliosis (AIS) may have concomitant pelvic height difference (PHD). There is limited data on PHD’s effect on surgical outcomes. We sought to evaluate if patients’ PHDs were associated with surgical outcomes in lowest instrumented vertebra translation (LIVT) and lowest instrumented vertebra angulation (LIVA) following spinal fusion.

Methods: We performed a single-center retrospective cohort study. Inclusion criteria consisted of patients who underwent fusion for AIS from 2006 – 2018 and had imaging at one-year follow-up. All patients treated for non-idiopathic scoliosis or who lacked imaging were excluded.

Demographic data was collected, and patients were grouped based on their preoperative PHD (0-5mm, 6-10mm, and 10+mm). Using a general linear model, we analyzed differences between PHD groups in improvement in LIVT and LIVA at one year when accounting for preoperative values. Assumptions of normality, homogeneity of variance-covariance, and linearity were satisfied.

Wilk’s Lambda multivariate test was used to determine overall differences between groups. Following multivariate analysis, the test of between-subjects effects was used to determine univariate significance and effect power of group membership on each dependent variable. Next, a pairwise comparison was done to demonstrate significance and effect power between groups. Lastly, estimated marginal means were calculated to demonstrate differences in surgical outcomes by group while maintaining preoperative LIVT and LIVA values at the mean.

Results: Two hundred and twenty-nine patients met the inclusion criteria. The mean age was 14.2 years and 87% were female. There were 140, 47, and 42 patients in the 0-5mm, 6-10mm, and 10+mm groups, respectively.

LIVT and LIVA correction were different among PHD groups (Wilk’s Λ = .958, F(4, 446) = 2.447, p = 0.046, partial η2 = 0.021) on multivariate analysis. Preoperative PHD was associated with differences in improvement in LIVT (p = 0.01; partial η2 = 0.04) but not LIVA (p = 0.535; partial η2 = 0.006) on univariate between-subjects analysis. There was a significant difference found between patients with 0-5mm and 10+mm PHD (p = 0.003) and 6-10mm and 10+mm (p = 0.030) in improvement in LIVT only on pairwise comparison.

When examining the estimated marginal means of improvement in LIVT and LIVA at the mean values of preoperative LIVT (17.90mm) and LIVA (16.94 degrees), we see significant differences between the 0-5mm (8.009; 95% CI, 6.881 - 9.138) and 10+mm (4.255; 95% CI, 2.142 - 6.368) groups for improvement in LIVT, but not LIVA.

Conclusion: Patients with different preoperative PHD were significantly associated with different surgical outcomes in improvement in LIVT but not LIVA following spinal fusion when accounting for preoperative LIVT and LIVA. As indicated by multivariate effect size (partial η2 = 0.021), univariate effect size (partial η2 = 0.04), and estimated marginal means, however, this effect is small.