1
|
Guo Z, Wang P, Ye S, Li H, Bao J, Shi R, Yang S, Yin R, Wu X. Interpretable Machine Learning Models Based on Shapley Additive Explanations for Predicting the Risk of Cerebrospinal Fluid Leakage in Lumbar Fusion Surgery. Spine (Phila Pa 1976) 2024; 49:1281-1293. [PMID: 38963261 DOI: 10.1097/brs.0000000000005087] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/07/2024] [Accepted: 06/17/2024] [Indexed: 07/05/2024]
Abstract
STUDY DESIGN Retrospective study. OBJECTIVES The objective of this investigation was to formulate and internally verify a customized machine learning (ML) framework for forecasting cerebrospinal fluid leakage (CSFL) in lumbar fusion surgery. This was accomplished by integrating imaging parameters and using the SHapley Additive exPlanation (SHAP) technique to elucidate the interpretability of the model. SUMMARY OF BACKGROUND DATA Given the increasing incidence and surgical volume of spinal degeneration worldwide, accurate predictions of postoperative complications are urgently needed. SHAP-based interpretable ML models have not been used for CSFL risk factor analysis in lumbar fusion surgery. METHODS Clinical and imaging data were retrospectively collected from 3505 patients who underwent lumbar fusion surgery. Six distinct machine learning models were formulated: extreme gradient boosting (XGBoost), decision tree (DT), random forest (RF), support vector machine (SVM), Gaussian naive Bayes (GaussianNB), and K-nearest neighbors (KNN) models. Evaluation of model performance on the test dataset was performed using performance metrics, and the analysis was executed through the SHAP framework. RESULTS CSFL was detected in 95 (2.71%) of 3505 patients. Notably, the XGBoost model exhibited outstanding accuracy in forecasting CSFLs, with high precision (0.9815), recall (0.6667), accuracy (0.8182), F1 score (0.7347), and AUC (0.7343). In addition, through SHAP analysis, significant predictors of CSFL were identified, including ligamentum flavum thickness, zygapophysial joint degeneration grade, central spinal stenosis grade, decompression segment count, decompression mode, intervertebral height difference, Cobb angle, intervertebral height index difference, operation mode, lumbar segment lordosis angle difference, Meyerding grade of lumbar spondylolisthesis, and revision surgery. CONCLUSIONS The combination of the XGBoost model with the SHAP is an effective tool for predicting the risk of CSFL during lumbar fusion surgery. Its implementation could aid clinicians in making informed decisions, potentially enhancing patient outcomes and lowering healthcare expenses. This study advocates for the adoption of this approach in clinical settings to enhance the evaluation of CSFL risk among patients undergoing lumbar fusion.
Collapse
Affiliation(s)
- ZongJie Guo
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - PeiYang Wang
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - SuHui Ye
- Department of Orthopedics, Yancheng Third People's Hospital, Yancheng, Jiangsu, People's Republic of China
| | - HaoYu Li
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - JunPing Bao
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - Rui Shi
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - Shu Yang
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| | - Rui Yin
- Department of immunology, Institute of Translational Medicine, Medical College, Yangzhou University, Yangzhou, Jiangsu, People's Republic of China
| | - XiaoTao Wu
- Spine Surgery Center, Department of Spine Surgery, Zhongda Hospital Affiliated to Southeast University, Nanjing, Jiangsu, People's Republic of China
| |
Collapse
|
2
|
Stoltzfus MT, Nguyen K, Freedman Z, Hallan DR, Hong J, Rizk E. Effects of Preoperative Glucocorticoid Use on Patients Undergoing Single-Level Lumbar Fusions: A Retrospective Propensity Score-Matched Registry Study. Cureus 2024; 16:e57197. [PMID: 38681464 PMCID: PMC11056191 DOI: 10.7759/cureus.57197] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/14/2024] [Accepted: 03/26/2024] [Indexed: 05/01/2024] Open
Abstract
Objective Spinal fusions are gaining popularity as a means of treating spinal deformity and instability from a range of pathologies. The prevalence of glucocorticoid use has also increased in recent decades, and their systemic effects are well-documented. Although commonly used in the preoperative period, the effects of steroids on outcomes among patients undergoing spinal fusions are inadequately described. This study compares the odds of developing complications among patients who underwent single-level lumbar fusions with and without preoperative glucocorticoid use in hopes of establishing more evidence-based parameters for guiding preoperative steroid use. Methods The TriNetX multi-institutional electronic health record database was used to perform a retrospective, propensity score-matched analysis of clinical outcomes of two cohorts of patients who underwent posterior or posterolateral single-level lumbar fusions with and without interbody fusion, those who used glucocorticoids for at least one week within a year of fusion and those who did not. The outcomes of interest were examined within 30 days of the operation and included death, reoperation, deep or superficial surgical site infection (SSI), pneumonia, reintubation, ventilator dependence, tracheostomy, acute kidney injury (AKI), renal insufficiency, pulmonary embolism (PE) or deep venous thrombosis (DVT), urinary tract infection (UTI), emergency department (ED) visit, sepsis, and myocardial infarction (MI). Results The odds of developing pneumonia within 30 days of spinal fusion in the cohort that used glucocorticoids within one year of operation compared to the cohort without glucocorticoid use was 0.67 (p≤0.001, 95% CI: 0.59-0.69). The odds of requiring a tracheostomy within 30 days of spinal fusion in the cohort that used glucocorticoids within one year of operation compared to the cohort without glucocorticoid use was 0.39 (p≤0.001, 95% CI: 0.26-0.60). The odds of reoperation, deep and superficial SSI, and ED visits within 30 days of operation were significantly higher for the same glucocorticoid-receiving cohort, with odds ratios of 1.4 (p=0.003, 95% CI: 1.11-1.65), 1.86 (p≤0.001, 95% CI: 1.31-2.63), 2.28 (p≤0.001, 95% CI: 1.57-3.31), and 1.25 (p≤0.001, 95% CI: 1.17-1.33), respectively. After propensity score-matching, there was no significant difference between the odds of death, DVT, PE, MI, UTI, AKI, sepsis, reintubation, and ventilator dependence between the two cohorts. Conclusion In support of much of the current literature regarding preoperative glucocorticoid use and rates of complications, patients who underwent a single-level lumbar fusion and have used glucocorticoids for at least a week within a year of operation experienced significantly higher odds of reoperation, deep and superficial SSI, and ED visits. However, these patients using glucocorticoids were also found to have lower odds of developing pneumonia, renal insufficiency, and tracheostomy requirement than those who did not use steroids within a year of surgery.
Collapse
Affiliation(s)
- Mason T Stoltzfus
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| | - Kenny Nguyen
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| | - Zachary Freedman
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| | - David R Hallan
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| | - Jinpyo Hong
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| | - Elias Rizk
- Department of Neurosurgery, Penn State University College of Medicine, Milton S. Hershey Medical Center, Hershey, USA
| |
Collapse
|
3
|
Gong JH, Sastry R, Koh DJ, Soliman L, Sobti N, Oyelese AA, Gokaslan ZL, Fridley J, Woo AS. Early Outcomes of Muscle Flap Closures in Posterior Thoracolumbar Fusions: A Propensity-Matched Cohort Analysis. World Neurosurg 2023; 180:e392-e407. [PMID: 37769839 DOI: 10.1016/j.wneu.2023.09.078] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/23/2023] [Revised: 09/18/2023] [Accepted: 09/20/2023] [Indexed: 10/03/2023]
Abstract
BACKGROUND Plastic surgery closure with muscle flaps after complex spinal reconstruction has become increasingly common. Existing evidence for this practice consists of small, uncontrolled, single-center cohort studies. We aimed to compare 30-day postoperative wound-related complication rates between flap closure and traditional closure after posterior thoracolumbar fusions (PTLFs) for non-infectious, non-oncologic pathologies using a national database. METHODS We performed a propensity-matched analysis using the 2012-2020 National Surgical Quality Improvement Program dataset to compare 30-day outcomes between PTLFs with flap closure versus traditional closure. RESULTS A total of 100,799 PTLFs met our inclusion criteria. The use of flap closure with PTLF remained low but more than doubled from 2012 to 2020 (0.38% vs. 0.97%; P = 0.002). A higher proportion of flap closures had higher American Society of Anesthesiologists classifications and higher number of operated spine levels (all P < 0.001). We included 1907 PTLFs (630 for flap closure; 1257 for traditional closure) in the propensity-matched cohort. Unadjusted 30-day wound complication rates were 1.7% for flap and 2.1% for traditional closure (P = 0.76). After adjusting for operative time, wound complication, readmission, reoperation, mortality, and non-wound complication were not associated flap use (all P > 0.05). CONCLUSIONS Plastic surgery closure was performed in patients with a higher comorbidity burden, suggesting consultation in sicker patients. Although higher rates of wound and non-wound complications were expected for the flap cohort, our propensity-matched cohort analysis of flap closure in PTLFs resulted in non-inferior odds of wound complications compared to traditional closure. Further study is needed to assess long-term complications in prophylactic flap closure in complex spine surgeries.
Collapse
Affiliation(s)
- Jung Ho Gong
- Division of Plastic and Reconstructive Surgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA.
| | - Rahul Sastry
- Department of Neurosurgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Daniel J Koh
- Boston University School of Medicine, Boston, Massachusetts, USA
| | - Luke Soliman
- Division of Plastic and Reconstructive Surgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Nikhil Sobti
- Division of Plastic and Reconstructive Surgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Adetokunbo A Oyelese
- Department of Neurosurgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Ziya L Gokaslan
- Department of Neurosurgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Jared Fridley
- Department of Neurosurgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| | - Albert S Woo
- Division of Plastic and Reconstructive Surgery, The Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
| |
Collapse
|
4
|
Hung YT, Hung WK, Chi CC. Effects of Preoperative Chronic Steroid Use on Postoperative Outcomes in Orthopedic Surgery: A Systematic Review and Meta-Analysis. Pharmaceuticals (Basel) 2023; 16:1328. [PMID: 37765136 PMCID: PMC10536822 DOI: 10.3390/ph16091328] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/15/2023] [Revised: 09/17/2023] [Accepted: 09/18/2023] [Indexed: 09/29/2023] Open
Abstract
Higher rates of postoperative complications have been found in preoperative chronic steroid users. However, the effects of preoperative chronic steroid use on outcomes in orthopedic surgery were unclear. We performed a systematic review of cohort studies examining the effects of chronic steroid use on postoperative outcomes following orthopedic surgery and searched PubMed, Embase, and CENTRAL through 29 April 2023. We included 17 studies with 1,546,562 patients. No increase in 30-day mortality (adjusted odds ratio (aOR) 1.40, 95% confidence interval (CI) 0.64-3.09) and composite thromboembolic events (aOR 1.61, 95% CI 0.99-2.63) but increases in 30-day overall complications (aOR 1.42, 95% CI 1.16-1.75), wound dehiscence (aOR 2.91, 95% CI 1.49-5.66), infectious complications (any infection (aOR 1.61, 95% CI 1.44-1.80), sepsis (aOR 2.07, 95% CI 1.34-3.21), superficial surgical site infection (SSI) (aOR 1.73, 95% CI 1.03-2.89) and deep SSI (aOR 1.96, 95% CI 1.26-3.05)), re-admission (aOR 1.62, 95% CI 1.48-1.77), both 30-day (aOR 1.28, 95% CI 1.03-1.59) and 1-year re-operation (aOR 1.78, 95% CI 1.09-2.92), pulmonary embolism (aOR 5.94, 95% CI 1.52-23.29), and deep vein thrombosis (aOR 2.07, 95% CI 1.24-3.46) were detected in preoperative steroid users. An increased risk of adverse outcomes following orthopedic surgery in chronic steroid users was found.
Collapse
Affiliation(s)
- Yu-Ting Hung
- Department of Anesthesiology, Chang Gung Memorial Hospital, Linkou, Taoyuan 33305, Taiwan;
| | - Wei-Kai Hung
- Department of Dermatology, Chang Gung Memorial Hospital, Linkou, 5, Fuxing St., Guishan Dist., Taoyuan 33305, Taiwan;
| | - Ching-Chi Chi
- Department of Dermatology, Chang Gung Memorial Hospital, Linkou, 5, Fuxing St., Guishan Dist., Taoyuan 33305, Taiwan;
- School of Medicine, College of Medicine, Chang Gung University, Taoyuan 33302, Taiwan
| |
Collapse
|
5
|
Ifarraguerri AM, Gupta P, Quan T, Cohen JS, Chen FR, Zeitlin J, Manzi JE, Farley B, Ramamurti P, Tabaie S. Risks of Immunosuppressive Therapy in Patients Undergoing Open Reduction Internal Fixation for Ankle Fractures. J Foot Ankle Surg 2023; 62:802-806. [PMID: 37086904 DOI: 10.1053/j.jfas.2023.04.003] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/22/2022] [Revised: 04/04/2023] [Accepted: 04/12/2023] [Indexed: 04/24/2023]
Abstract
Chronic steroid and immunosuppressant use have been shown to increase the risk for postoperative complications in orthopedic surgery. Further understanding of the risks of immunosuppression is necessary to aid in risk stratification and patient counseling. However, these risks have not yet been explored in ankle fracture patients. Thus, the purpose of this study is to determine whether patients taking immunosuppressives are at an increased risk for morbidity and mortality following open reduction and internal fixation (ORIF) of ankle fractures. Patients undergoing operative treatment for ankle fractures from 2006 to 2018 were identified in the National Surgical Quality Improvement Program database. Patients were categorized based on their use of immunosuppressive medications. Postoperative outcomes assessed included superficial surgical site infections, deep surgical site infections, organ space infections, wound dehiscence, pneumonia, unplanned intubation, pulmonary embolism, urinary tract infection, renal failure, blood transfusion requirement, deep vein thrombosis, sepsis, cardiac arrest, extended length of hospital stay, readmission, reoperation, and mortality. Univariate and multivariate analyses were performed. In total, 10,331 patients underwent operative treatment for ankle fracture. Total 10,153 patients (98.3%) were not taking immunosuppressants and 178 (1.7%) were taking these medications. In multivariate analysis, patients taking immunosuppressants were at increased risk of pulmonary embolism (odds ratio [OR] 4.382; p = .041) and hospital readmission (OR 2.131; p = .021). Use of immunosuppressive medications is an independent risk factor for pulmonary embolism and readmission following ORIF for ankle fractures. Notably, no association with wound complications, infections, or sepsis was identified.
Collapse
Affiliation(s)
- Anna M Ifarraguerri
- Department of Orthopaedic Surgery, The George Washington School of Medicine and Health Sciences, Washington, DC
| | - Puneet Gupta
- Department of Orthopaedic Surgery, The George Washington School of Medicine and Health Sciences, Washington, DC
| | - Theodore Quan
- Department of Orthopaedic Surgery, The George Washington School of Medicine and Health Sciences, Washington, DC.
| | - Jordan S Cohen
- Department of Orthopaedic Surgery, University of Pennsylvania, Philadelphia, PA
| | - Frank R Chen
- Department of Anesthesiology, Hospital of the University of Pennsylvania, Philadelphia, PA
| | | | | | - Benjamin Farley
- Department of Orthopaedic Surgery, The George Washington School of Medicine and Health Sciences, Washington, DC
| | - Pradip Ramamurti
- Department of Orthopaedic Surgery, University of Virginia Medical Center, Charlottesville, VA
| | - Sean Tabaie
- Department of Orthopaedic Surgery, Children's National Health System, Washington, DC
| |
Collapse
|
6
|
Liu Z, Tang Q, Liu RT, Yu MZ, Peng H, Zhang CQ, Zhu ZZ, Wei XJ. Laponite intercalated biomimetic multilayer coating prevents glucocorticoids induced orthopedic implant failure. Bioact Mater 2023; 22:60-73. [PMID: 36203962 PMCID: PMC9519439 DOI: 10.1016/j.bioactmat.2022.09.013] [Citation(s) in RCA: 7] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/11/2022] [Revised: 08/18/2022] [Accepted: 09/12/2022] [Indexed: 11/25/2022] Open
|