Recovery kinetics following spinal deformity correction: a comparison of isolated cervical, thoracolumbar, and combined deformity morphometries.

BACKGROUND CONTEXT: The postoperative recovery patterns of cervical deformity patients, thoracolumbar deformity patients, and patients with combined cervical and thoracolumbar deformities, all relative to one another, is not well understood. Clear objective benchmarks are needed to quantitatively define a "good" versus a "bad" postoperative recovery across multiple follow-up visits, varying deformity types, and guide expectations. PURPOSE: To objectively define and compare the complete 2-year postoperative recovery process among operative cervical only, thoracolumbar only, and combined deformity patients using area-under-the-curve (AUC) methodology. STUDY DESIGN/SETTING: Retrospective review of 2 prospective, multicenter adult cervical and spinal deformity databases. PATIENT SAMPLE: One hundred seventy spinal deformity patients. OUTCOME MEASURES: Common health-related quality of life (HRQOL) assessments across both databases included the EuroQol 5-Dimension Questionnaire and Numeric Rating Scale (NRS) back pain assessment. In order to compare disability improvements, the Neck Disability Index (NDI) and the Oswestry Disability Index (ODI) were merged into one outcome variable, the ODI-NDI. Both assessments are gauged on the same scale, with minimal question deviation. Sagittal Radiographic Alignment was also assessed at pre- and all postoperative time points. METHODS: Operative deformity patients >18 years old with baseline (BL) to 2-year HRQOLs were included. Patients were stratified by cervical only (C), thoracolumbar only (T), and combined deformities (CT). HRQOL and radiographic outcomes were compared within and between deformity groups. AUC normalization generated normalized HRQOL scores at BL and all follow-up intervals (6 weeks, 3 months, 1 year, and 2 year). Normalized scores were plotted against follow-up time interval. AUC was calculated for each follow-up interval, and total area was divided by cumulative follow-up length, determining overall, time-adjusted HRQOL recovery (Integrated Health State, IHS). Multiple linear regression models determined significant predictors of HRQOL discrepancies among deformity groups. RESULTS: One hundred seventy patients were included (27 C, 27 T, and 116 CT). Age, BMI, sex, smoking status, osteoporosis, depression, and BL HRQOL scores were similar among groups (p >. 05). T and CT patients had higher comorbidity severities (CCI: C 0.696, T 1.815, CT 1.699, p = .020). Posterior surgical approaches were most common (62.9%) followed by combined (28.8%) and anterior (6.5%). Standard HRQOL analysis found no significant differences among groups until 1-year follow-up, where C patients exhibited comparatively greater NRS back pain (4.88 vs. 3.65 vs. 3.28, p = .028). NRS Back pain differences between groups subsided by 2-years (p>.05). Despite C patients exhibiting significantly faster ODI-NDI minimal clinically important difference (MCID) achievement (33.3% vs. 0% vs. 23.0%, p < .001), all deformity groups exhibited similar ODI-NDI MCID achievement by 2-years (51.9% vs. 59.3% vs. 62.9%, p = 0.563). After HRQOL normalization, similar results were observed relative to the standard analysis (1-year NRS Back: C 1.17 vs. T 0.50 vs. CT 0.51, p < .001; 2-year NRS Back: 1.20 vs. 0.51 vs. 0.69, p = .060). C patients exhibited a worse NRS back normalized IHS (C 1.18 vs. T 0.58 vs. CT 0.63, p = .004), indicating C patients were in a greater state of postoperative back pain for a longer amount of time. Linear regression models determined postoperative distal junctional kyphosis (adjusted beta: 0.207, p = .039) and osteoporosis (adjusted beta: 0.269, p = .007) as the strongest predictors of a poor NRS back IHS (model summary: R2 = 0.177, p = .039). CONCLUSIONS: Despite C patients exhibiting a quicker rate of MCID disability (ODI-NDI) improvement, they exhibited a poorer overall recovery of back pain with worse NRS back scores compared with BL status and other deformity groups. Postoperative distal junctional kyphosis and osteoporosis were identified as primary drivers of a poor postoperative NRS back IHS. Utilization of the IHS, a single number adjusting for all postoperative HRQOL visits, in conjunction with predictive modelling may pose as an improved method of gauging the effect of surgical details and complications on a patient's entire recovery process.

Recovery Kinetics: Comparison of Patients Undergoing Primary or Revision Procedures for Adult Cervical Deformity Using a Novel Area Under the Curve Methodology.

BACKGROUND: Limited data are available to objectively define what constitutes a "good" versus a "bad" recovery for operative cervical deformity (CD) patients. Furthermore, the recovery patterns of primary versus revision procedures for CD is poorly understood. OBJECTIVE: To define and compare the recovery profiles of CD patients undergoing primary or revision procedures, utilizing a novel area-under-the-curve normalization methodology. METHODS: CD patients undergoing primary or revision surgery with baseline to 1-yr health-related quality of life (HRQL) scores were included. Clinical symptoms and HRQL were compared among groups (primary/revision). Normalized HRQL scores at baseline and follow-up intervals (3M, 6M, 1Y) were generated. Normalized HRQLs were plotted and area under the curve was calculated, generating one number describing overall recovery (Integrated Health State). Subanalysis identified recovery patterns through 2-yr follow-up. RESULTS: Eighty-three patients were included (45 primary, 38 revision). Age (61.3 vs 61.9), gender (F: 66.7% vs 63.2%), body mass index (27.7 vs 29.3), Charlson Comorbidity Index, frailty, and osteoporosis (20% vs 13.2%) were similar between groups (P > .05). Primary patients were more preoperatively neurologically symptomatic (55.6% vs 31.6%), less sagittally malaligned (cervical sagittal vertical axis [cSVA]: 32.6 vs 46.6; T1 slope: 28.8 vs 36.8), underwent more anterior-only approaches (28.9% vs 7.9%), and less posterior-only approaches (37.8% vs 60.5%), all P < .05. Combined approaches, decompressions, osteotomies, and construct length were similar between groups (P > .05). Revisions had longer op-times (438.0 vs 734.4 min, P = .008). Following surgery, complication rate was similar between groups (66.6% vs 65.8%, P = .569). Revision patients remained more malaligned (cSVA, TS-CL; P < .05) than primary patients until 1-yr follow-up (P > .05). Normalized HRQLs determined primary patients to exhibit less neck pain (numeric rating scale [NRS]) and myelopathy (modified Japanese Orthopaedic Association) symptoms through 1-yr follow-up compared to revision patients (P < .05). These differences subsided when following patients through 2 yr (P > .05). Despite similar 2-yr HRQL outcomes, revision patients exhibited worse neck pain (NRS) Integrated Health State recovery (P < .05). CONCLUSION: Despite both primary and revision patients exhibiting similar HRQL outcomes at final follow-up, revision patients were in a greater state of postoperative neck pain for a greater amount of time.