Supporting data for “Postoperative Neurological Complications in Spine Surgery: Identifying Risk Factors and Predictive Modelling”

<p dir="ltr">The thesis investigates the risk factors (Chapter 2) and the development of predictive models (Chapter 3) for postoperative neurological complications in spine surgery. The study designs for Chapter 2 and Chapter 3 are a systematic review with meta-analysis and a retrospective study respectively. To support this work, the dataset submission includes processed and raw data, as well as the related figures (Figures 1-24 for Chapter 2 and Figures 25-33 for Chapter 3).</p><p dir="ltr">The details of the figures are presented as follows:</p><p dir="ltr">1) For "Figures of Section 2.3" in Chapter 2 file </p><p dir="ltr">-Figure 1. Flow diagram detailing study selection process</p><p dir="ltr">-Figure 2. Forest chart of the overall pooled incidence estimates of postoperative neurological complications in spine surgery</p><p dir="ltr">-Figure 3. Forest plot of pooled incidence estimates of postoperative neurological complications in spine surgery by study region subgroups</p><p dir="ltr">-Figure 4. Forest plot of pooled incidence estimates of postoperative neurological complications in spine surgery by gender subgroups</p><p dir="ltr">-Figure 5. Forest plot of pooled incidence estimates of postoperative neurological complications in spine surgery by type of spinal disease subgroups</p><p dir="ltr">-Figure 6. Forest plot representing odds ratios of postoperative neurological complications in the male gender </p><p dir="ltr">-Figure 7. Forest plot representing odds ratios of postoperative neurological complications associated with age-related factors</p><p dir="ltr">-Figure 8. Forest plot representing odds ratios of postoperative neurological complications associated with diabetes mellitus</p><p dir="ltr">-Figure 9. Forest plot representing odds ratios of postoperative neurological complications in pedicle subtraction osteotomy</p><p dir="ltr">-Figure 10. Bubble plot of meta regression analysis representing postoperative neurological complications against sample size</p><p dir="ltr">-Figure 11. Leave-one-out analysis of the included studies</p><p dir="ltr">2) For "Figures of Section 2.4" in Chapter 2 file</p><p dir="ltr">-Figure 12. Flow diagram detailing study selection process</p><p dir="ltr">-Figure 13. Publication bias of all included studies</p><p dir="ltr">-Figure 14. Forest plot of pooled incidence estimates of postoperative neurological complications in spinal deformity surgery</p><p dir="ltr">-Figure 15. Forest plot of pooled incidence estimates of postoperative neurological complications in spinal deformity surgery by type of spinal deformity subgroup</p><p dir="ltr">-Figure 16. Forest plot of pooled incidence estimates of postoperative neurological complications in spinal deformity surgery by surgical procedure subgroup</p><p dir="ltr">-Figure 17. Forest plot representing odds ratios of postoperative neurological complications for age at surgery</p><p dir="ltr">-Figure 18. Forest plot representing odds ratios of postoperative neurological complications for BMI</p><p dir="ltr">-Figure 19. Forest plot representing odds ratios of postoperative neurological complications for diabetes mellitus</p><p dir="ltr">-Figure 20. Forest plot representing odds ratios of postoperative neurological complications</p><p dir="ltr">for the presence of kyphosis</p><p dir="ltr">-Figure 21. Forest plot representing odds ratios of postoperative neurological complications for blood loss</p><p dir="ltr">-Figure 22. Forest plot representing odds ratios of postoperative neurological complications for preoperative neurological deficits</p><p dir="ltr">-Figure 23. Bubble plot of meta-regression analysis representing postoperative neurological complications against sample size</p><p dir="ltr">-Figure 24. Leave-one-out analysis of the included studies</p><p><br></p><p dir="ltr">3) For "Figures" in Chapter 3 file</p><p dir="ltr">-Figure 25. Flow diagram detailing study selection process</p><p dir="ltr">-Figure 26. Predictor selection using LASSO logistic regression analysis</p><p dir="ltr">-Figure 27. The importance ranking of study variables identified by Random Forest analysis</p><p dir="ltr">-Figure 28. The importance ranking of study variables identified by XGBoost analysis</p><p dir="ltr">-Figure 29. Heatmap presenting the relationships between selected study variables with MIC values</p><p dir="ltr">-Figure 30. A nomogram for predicting the risk of postoperative neurologic deficits</p><p dir="ltr">-Figure 31. ROC curves of the predictive model</p><p dir="ltr">-Figure 32. Calibration curves of the predictive model</p><p dir="ltr">-Figure 33. DCA of the predictive model</p><p dir="ltr">The processed data for data analysis for Section 2.3 and Section 2.4 of the thesis are presented in Excel spreasheet respectively.</p><p dir="ltr">For "processed data for section 2.3" in Chapter 2 file, the author, publication year, case of neurological complications, sample size and study region have been shown.</p><p dir="ltr">For “processed data for section 2.4” in Chapter 2 file, the author, publication year, case of neurological complications, sample size have been shown.</p><p dir="ltr">The processed data used for analyses in Sections 2.3 and 2.4 of the thesis are provided in separate Excel spreadsheets.</p><p dir="ltr">For the “processed data for Section 2.3” in the Chapter 2 file, the data include author, year of publication, number of neurological complication cases, sample size and study region.</p><p dir="ltr">For the “processed data for Section 2.4” in the Chapter 2 file, the data include author, year of publication, number of neurological complication cases and sample size.</p><p dir="ltr">In the “recruited patients (n=249)” in the Chapter 3 file, all data have been anonymized to ensure the removal of confidential information in accordance with IRB regulations.</p><p dir="ltr">Study variables include BMI, comorbidities such as diabetes mellitus, hypertension, hyperlipidaemia, renal impairment or chronic kidney disease, cerebrovascular disease, depression or anxiety and preoperative pain status. Lifestyle risk factors, including smoking and alcohol consumption, were also recorded. Data on anticoagulation and antiplatelet therapy were extracted.</p><p dir="ltr">Surgical and intraoperative variables included timing of surgery, type of surgery, surgical approach, patient position, number of vertebral levels operated, surgery duration, anesthesia duration, and intraoperative blood loss. Additional factors assessed included surgeon experience, number of levels (single vs. multiple), and use of IONM. Intraoperative details included whether minimally invasive techniques were employed, changes in patient position, bone removal techniques, bone graft use and the need for dural repair. The use of haemostatic agents, antibiotics, inhalation anesthetic agents, the Valsalva manoeuvre and intraoperative transfusion were also documented, along with operation class, American Society of Anesthesiologists (ASA) status, surgical diagnosis and surgical segment.</p><p><br></p>