P95. Geographic Variation in the Philosophy of Adult Spinal Deformity Management and Patient Outcomes: A Propensity Score Matching Analysis from the Prospective Evaluation Deformity Surgery (PEEDS) Study
Kenny Kwan,Stephen J. Lewis,Lawrence G. Lenke,Christopher I. Shaffrey,Sigurd H. Berven,Yong Qiu,Yukihiro Matsuyama,Ferran Pellise,David W. Polly,Jonathan N. Sembrano,Michael P. Kelly,Benny Dahl,Maarten Spruit,Ahmet Alanay,Justin S. Smith,Kenneth M. Cheung
DOI: https://doi.org/10.1016/j.spinee.2021.05.303
IF: 4.297
2021-01-01
The Spine Journal
Abstract:BACKGROUND CONTEXT Severity of deformity, preoperative disability, and surgical strategies are variable between, and within, centers performing reconstructive surgery for spinal deformity. However, geographic variation in how surgeons manage ASD patients worldwide is unknown. PURPOSE The purpose of this study was to compare the surgical approach and clinical outcomes of elderly patients treated with fusion for spinal deformity in North America and Asia. STUDY DESIGN/SETTING Retrospective review of the prospectively collected data. PATIENT SAMPLE All patients enrolled in the database were eligible and were propensity score matched according to their region (North America and Asia) OUTCOME MEASURES Surgical details, postoperative radiological parameter and health-related quality of life (HRQoL) scores at preoperative and 2 years, and occurrence of adverse events within the study period were compared in the matched pairs with dependent statistical tests. METHODS The two largest regions (North America and Asia) from the Prospective Evaluation Deformity Surgery (PEEDS) database, consisting of ASD patients over aged 60 with primary spinal fusion of at least 5 segments with a minimum of 2 year follow-up, were compared. Patients were 1:1 propensity score-matched using age, gender, Charlson Comorbidity Index, lumbar lordosis (LL), pelvic incidence (PI)-LL mismatch, sagittal vertical axis (SVA), absolute lumbar cobb angle, and sacral slope for calculation of propensity score and applying greedy algorithm to identify the best matches between the 2 regions with maximal absolute distance of 0.25 in propensity score to each other. RESULTS Out of the 219 patients enrolled in the database, 82 propensity score-matched patients were identified (n=41 in each region), with a median age of 68.0 (range 64-72) years (NA) and 68.0 (range 63-73) years (A). Both cohorts had comparable distributions of all preoperative radiological parameters and HRQoL scores. Out of the 219 patients enrolled in the database, 82 propensity score-matched patients were identified (n=41 in each region), with a median age of 68.0 (range 64-72) years (NA) and 68.0 (range 63-73) years (A). Both cohorts had comparable distributions of all preoperative radiological parameters and HRQoL scores. Patients treated in NA had higher proportion of single stage surgery (88% vs 44%, p<0.001), more levels fused throughout all stages (11.5±3.7 vs 9.1±2.3, p<0.001), and higher cumulative median blood loss (1500ml vs 1200ml, p=0.010) compared with those treated in A. Surgical procedures differed between the two groups: posterior instrumentation with TLIF were predominant in NA (83%), whereas a combined lateral and posterior approach (49% vs 5%) was more common in A. The occurrence of having any adverse events was higher in the NA group (82% vs 35%, p=0.002), due to a difference in the incidence of deep vein thrombosis (21% vs 0%, p=0.016) and other AEs (44% vs 15%, p=0.013). At 2-year follow-up, there was no difference in the correction of radiological parameters between the two groups. In terms of HRQoL, at 2 year follow-up patients in NA had a more favourable SRS-22r total score, EQ-5D VAS and leg pain NRS compared with the matched pairs in A, but similar EQ-5D index, ODI and back pain scores. CONCLUSIONS In this propensity score matched analysis, we demonstrated marked geographic variation in surgical practice toward surgeries for adult spinal deformity despite similar baseline patient characteristics and radiological parameters between North America and Asia. However, the 2 year radiological and clinical outcomes were similar. Further study is necessary to understand the drivers of this variation. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs. Severity of deformity, preoperative disability, and surgical strategies are variable between, and within, centers performing reconstructive surgery for spinal deformity. However, geographic variation in how surgeons manage ASD patients worldwide is unknown. The purpose of this study was to compare the surgical approach and clinical outcomes of elderly patients treated with fusion for spinal deformity in North America and Asia. Retrospective review of the prospectively collected data. All patients enrolled in the database were eligible and were propensity score matched according to their region (North America and Asia) Surgical details, postoperative radiological parameter and health-related quality of life (HRQoL) scores at preoperative and 2 years, and occurrence of adverse events within the study period were compared in the matched pairs with dependent statistical tests. The two largest regions (North America and Asia) from the Prospective Evaluation Deformity Surgery (PEEDS) database, consisting of ASD patients over aged 60 with primary spinal fusion of at least 5 segments with a minimum of 2 year follow-up, were compared. Patients were 1:1 propensity score-matched using age, gender, Charlson Comorbidity Index, lumbar lordosis (LL), pelvic incidence (PI)-LL mismatch, sagittal vertical axis (SVA), absolute lumbar cobb angle, and sacral slope for calculation of propensity score and applying greedy algorithm to identify the best matches between the 2 regions with maximal absolute distance of 0.25 in propensity score to each other. Out of the 219 patients enrolled in the database, 82 propensity score-matched patients were identified (n=41 in each region), with a median age of 68.0 (range 64-72) years (NA) and 68.0 (range 63-73) years (A). Both cohorts had comparable distributions of all preoperative radiological parameters and HRQoL scores. Out of the 219 patients enrolled in the database, 82 propensity score-matched patients were identified (n=41 in each region), with a median age of 68.0 (range 64-72) years (NA) and 68.0 (range 63-73) years (A). Both cohorts had comparable distributions of all preoperative radiological parameters and HRQoL scores. Patients treated in NA had higher proportion of single stage surgery (88% vs 44%, p<0.001), more levels fused throughout all stages (11.5±3.7 vs 9.1±2.3, p<0.001), and higher cumulative median blood loss (1500ml vs 1200ml, p=0.010) compared with those treated in A. Surgical procedures differed between the two groups: posterior instrumentation with TLIF were predominant in NA (83%), whereas a combined lateral and posterior approach (49% vs 5%) was more common in A. The occurrence of having any adverse events was higher in the NA group (82% vs 35%, p=0.002), due to a difference in the incidence of deep vein thrombosis (21% vs 0%, p=0.016) and other AEs (44% vs 15%, p=0.013). At 2-year follow-up, there was no difference in the correction of radiological parameters between the two groups. In terms of HRQoL, at 2 year follow-up patients in NA had a more favourable SRS-22r total score, EQ-5D VAS and leg pain NRS compared with the matched pairs in A, but similar EQ-5D index, ODI and back pain scores. In this propensity score matched analysis, we demonstrated marked geographic variation in surgical practice toward surgeries for adult spinal deformity despite similar baseline patient characteristics and radiological parameters between North America and Asia. However, the 2 year radiological and clinical outcomes were similar. Further study is necessary to understand the drivers of this variation.