Predictive Modeling for Spinal Metastatic Disease.

Spinal metastasis is exceedingly common in patients with cancer and its prevalence is expected to increase. Surgical management of symptomatic spinal metastasis is indicated for pain relief, preservation or restoration of neurologic function, and mechanical stability. The overall prognosis is a major driver of treatment decisions; however, clinicians' ability to accurately predict survival is limited. In this narrative review, we first discuss the NOMS decision framework used to guide decision making in the treatment of patients with spinal metastasis. Given that decision making hinges on prognosis, multiple scoring systems have been developed over the last three decades to predict survival in patients with spinal metastasis; these systems have largely been developed using expert opinions or regression modeling. Although these tools have provided significant advances in our ability to predict prognosis, their utility is limited by the relative lack of patient-specific survival probability. Machine learning models have been developed in recent years to close this gap. Employing a greater number of features compared to models developed with conventional statistics, machine learning algorithms have been reported to predict 30-day, 6-week, 90-day, and 1-year mortality in spinal metastatic disease with excellent discrimination. These models are well calibrated and have been externally validated with domestic and international independent cohorts. Despite hypothesized and realized limitations, the role of machine learning methodology in predicting outcomes in spinal metastatic disease is likely to grow.

Similar Medium-Term Revision Rates Following Acute Total Hip Arthroplasty Versus Open Reduction and Internal Fixation for Acetabular Fractures in the Elderly.

BACKGROUND: The management of elderly acetabular fractures is complex, with high rates of conversion total hip arthroplasty (THA) after open reduction and internal fixation (ORIF), but potentially higher rates of complications after acute THA. METHODS: The California Office of Statewide Health Planning and Development database was queried between 2010 and 2017 for all patients aged 60 years or older who sustained a closed, isolated acetabular fracture and underwent ORIF, THA, or a combination. Chi-square tests and Student t tests were used to identify demographic differences between groups. Multivariate regression was used to evaluate predictors of 30-day readmission and 90-day complications. Kaplan-Meier (KM) survival analysis and Cox proportional hazards model were used to estimate the revision surgery-free survival (revision-free survival [RFS]), with revision surgery defined as conversion THA, revision ORIF, or revision THA. RESULTS: A total of 2,184 surgically managed acetabular fractures in elderly patients were identified, with 1,637 (75.0%) undergoing ORIF and 547 (25.0%) undergoing THA with or without ORIF. Median follow-up was 295 days (interquartile range, 13 to 1720 days). 99.4% of revisions following ORIF were for conversion arthroplasty. Unadjusted KM analysis showed no difference in RFS between ORIF and THA (log-rank test P = 0.27). RFS for ORIF patients was 95.1%, 85.8%, 78.3%, and 71.4% at 6, 12, 24 and 60 months, respectively. RFS for THA patients was 91.6%, 88.9%, 87.2%, and 78.8% at 6, 12, 24 and 60 months, respectively. Roughly 50% of revisions occurred within the first year postoperatively (49% for ORIF, 52% for THA). In propensity score-matched analysis, there was no difference between RFS on KM analysis (P = 0.22). CONCLUSIONS: No difference was observed in medium-term RFS between acute THA and ORIF for elderly acetabular fractures in California. Revision surgeries for either conversion or revision THA were relatively common in both groups, with roughly half of all revisions occurring within the first year postoperatively. LEVEL OF EVIDENCE: III.

A predictive algorithm for perioperative complications and readmission after ankle arthrodesis.

PURPOSE: Ankle arthrodesis is a mainstay of surgical management for ankle arthritis. Accurately risk-stratifying patients who undergo ankle arthrodesis would be of great utility. There is a paucity of accurate prediction models that can be used to pre-operatively risk-stratify patients for ankle arthrodesis. We aim to develop a predictive model for major perioperative complication or readmission after ankle arthrodesis. METHODS: This is a retrospective cohort study of adult patients who underwent ankle arthrodesis at any non-federal California hospital between 2015 and 2017. The primary outcome is readmission within 30 days or major perioperative complication. We build logistic regression and ML models spanning different classes of modeling approaches, assessing discrimination and calibration. We also rank the contribution of the included variables to model performance for prediction of adverse outcomes. RESULTS: A total of 1084 patients met inclusion criteria for this study. There were 131 patients with major complication or readmission (12.1%). The XGBoost algorithm demonstrates the highest discrimination with an area under the receiver operating characteristic curve of 0.707 and is well-calibrated. The features most important for prediction of adverse outcomes for the XGBoost model include: diabetes, peripheral vascular disease, teaching hospital status, morbid obesity, history of musculoskeletal infection, history of hip fracture, renal failure, implant complication, history of major fracture. CONCLUSION: We report a well-calibrated algorithm for prediction of major perioperative complications and 30-day readmission after ankle arthrodesis. This tool may help accurately risk-stratify patients and decrease likelihood of major complications.

Biportal Endoscopic Approach for Lumbar Degenerative Disease in the Ambulatory Outpatient vs Inpatient Setting: A Comparative Study.

BACKGROUND: Biportal spinal endoscopy is increasingly utilized for lumbar disc herniations and lumbar stenosis. The objective was to investigate the safety and effectiveness of the technique in the outpatient vs inpatient setting. METHODS: This is a comparative study of consecutive patients who underwent biportal spinal endoscopy by a single surgeon at a single institution. Demographics, surgical complications, and patient-reported outcomes were prospectively collected and retrospectively analyzed. Statistics were calculated among treatment groups using unpaired t test and χ 2 analysis where appropriate. Statistical significance was determined as P < 0.05. RESULTS: Eighty-four patients were included, 58 (69.0%) as outpatient, 26 (31.0%) as inpatient. Mean follow-up was 7.5 months. Statistically significant differences in age, American Society of Anesthesiologists classification, and Charleston Comorbidity Index scores were reported between cohorts, with younger and healthier patients undergoing outpatient surgery (P < 0.0001). Outpatients were more likely to have discectomies while inpatients were more likely to have decompressions for stenosis. No significant differences in postoperative complications were found between groups.Both cohorts demonstrated significant improvement in visual analog scale (VAS) back and leg pain scores and Oswestry Disability Index scores (P < 0.001). Outpatients had significantly lower postoperative VAS back pain (P = 0.001) and Oswestry Disability Index scores (P = 0.004) at 5-8 weeks compared with inpatients, but there was no significant difference for VAS leg pain scores at all time points between the cohorts. CONCLUSIONS: Early results demonstrate that biportal spinal endoscopy can safely and effectively be performed in both inpatient and outpatient settings. CLINICAL RELEVANCE: Outpatient biportal spinal endoscopy can be performed successfully in well selected patients, which may reduce the financial burden of spine surgery to the U.S. healthcare system.

Development and External Validation of a Risk Calculator for Prediction of Major Complications and Readmission After Anterior Cervical Discectomy and Fusion.

STUDY DESIGN: A retrospective, case-control study. OBJECTIVE: We aim to build a risk calculator predicting major perioperative complications after anterior cervical fusion. In addition, we aim to externally validate this calculator with an institutional cohort of patients who underwent anterior cervical discectomy and fusion (ACDF). SUMMARY OF BACKGROUND DATA: The average age and proportion of patients with at least one comorbidity undergoing ACDF have increased in recent years. Given the increased morbidity and cost associated with perioperative complications and unplanned readmission, accurate risk stratification of patients undergoing ACDF is of great clinical utility. METHODS: This is a retrospective cohort study of adults who underwent anterior cervical fusion at any nonfederal California hospital between 2015 and 2017. The primary outcome was major perioperative complication or 30-day readmission. We built standard and ensemble machine learning models for risk prediction, assessing discrimination, and calibration. The best-performing model was validated on an external cohort comprised of consecutive adult patients who underwent ACDF at our institution between 2013 and 2020. RESULTS: A total of 23,184 patients were included in this study; there were 1886 cases of major complication or readmissions. The ensemble model was well calibrated and demonstrated an area under the receiver operating characteristic curve of 0.728. The variables most important for the ensemble model include male sex, medical comorbidities, history of complications, and teaching hospital status. The ensemble model was evaluated on the validation cohort (n=260) with an area under the receiver operating characteristic curve of 0.802. The ensemble algorithm was used to build a web-based risk calculator. CONCLUSION: We report derivation and external validation of an ensemble algorithm for prediction of major perioperative complications and 30-day readmission after anterior cervical fusion. This model has excellent discrimination and is well calibrated when tested on a contemporaneous external cohort of ACDF cases.

Higher Rates of Readmission After Acute Total Hip Arthroplasty Versus Open Reduction Internal Fixation for Elderly Acetabular Fractures, a National Study From 2010 to 2019.

OBJECTIVES: To evaluate the initial complications and short-term readmissions and reoperations after open reduction internal fixation (ORIF) versus acute total hip arthroplasty (THA) for elderly acetabular fractures. DESIGN: Retrospective database review. SETTING: All hospitalizations in the National Readmissions Database and National Inpatient Sample. PATIENTS/PARTICIPANTS: Patients 60 years of age or older with closed acetabular fractures managed surgically identified from the National Readmissions Database or National Inpatient Sample between 2010 and 2019. INTERVENTION: Acute THA with or without ORIF. MAIN OUTCOME MEASUREMENTS: 30-, 90-, and 180-day readmissions and reoperations and index hospitalization complications. RESULTS: An estimated 12,538 surgically managed acetabular fractures in elderly patients occurred nationally between 2010 and 2019, with 10,008 (79.8%) undergoing ORIF and 2529 (20.2%) undergoing THA. Length of stay was 1.7 days shorter ( P < 0.001) and probability of nonhome discharge was reduced (OR 0.68, P = 0.009) for THA patients than for ORIF patients. THA was associated with lower rates of pneumonia (4.6 vs. 9.1%, P < 0.001) and other respiratory complications (10.2 vs. 17.6%) when compared with ORIF. At 30 days, THA patients had higher rates of readmission (13.9 vs. 10.1%, P = 0.007), related readmission (5.4 vs. 1.2%, P < 0.001), readmission for dislocation (3.1 vs. 0.3%, P < 0.001), and reoperations (2.9 vs. 0.9%, P = 0.002). At 180 days, THA patients had higher rates of related readmission (10.1% vs. 3.9%, P < 0.001), readmission for dislocation (5.1% vs. 1.3%, P < 0.001), and readmission for SSI (3.4 vs. 0.8%, P = 0.005). CONCLUSIONS: Acute THA is associated with lower length of stay and certain index hospitalization complications, but higher rates of readmissions for related reasons and specifically for dislocation. LEVEL OF EVIDENCE: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

External validation of a predictive algorithm for in-hospital and 90-day mortality after spinal epidural abscess.

BACKGROUND CONTEXT: Mortality in patients with spinal epidural abscess (SEA) remains high. Accurate prediction of patient-specific prognosis in SEA can improve patient counseling as well as guide management decisions. There are no externally validated studies predicting short-term mortality in patients with SEA. PURPOSE: The purpose of this study was to externally validate the Skeletal Oncology Research Group (SORG) stochastic gradient boosting algorithm for prediction of in-hospital and 90-day postdischarge mortality in SEA. STUDY DESIGN/SETTING: Retrospective, case-control study at a tertiary care academic medical center from 2003 to 2021. PATIENT SAMPLE: Adult patients admitted for radiologically confirmed diagnosis of SEA who did not initiate treatment at an outside institution. OUTCOME MEASURES: In-hospital and 90-day postdischarge mortality. METHODS: We tested the SORG stochastic gradient boosting algorithm on an independent validation cohort. We assessed its performance with discrimination, calibration, decision curve analysis, and overall performance. RESULTS: A total of 212 patients met inclusion criteria, with a short-term mortality rate of 10.4%. The area under the receiver operating characteristic curve (AUROC) of the SORG algorithm when tested on the full validation cohort was 0.82, the calibration intercept was -0.08, the calibration slope was 0.96, and the Brier score was 0.09. CONCLUSIONS: With a contemporaneous and geographically distinct independent cohort, we report successful external validation of a machine learning algorithm for prediction of in-hospital and 90-day postdischarge mortality in SEA.

Reduced Costs, Complications, and Length of Stay After Arthroscopic Versus Open Irrigation and Débridement for Knee Septic Arthritis.

BACKGROUND: In the treatment of native knee bacterial septic arthritis, the optimal irrigation and débridement modality-arthroscopic versus open-is a matter of controversy. We aim to compare revision-free survival, complications, and resource utilization between these approaches. METHODS: The National Readmission Database was queried from 2016 to 2019 to identify patients using International Classification of Diseases, 10th revision, diagnostic and procedure codes. Days to revision irrigation and débridement (I&D), if any, were calculated for patients during index admission or subsequent readmissions. Multivariate regression was used for healthcare utilization analysis. Survival analysis was done using Kaplan-Meier analysis and Cox proportional hazard regression. RESULTS: A total of 14,365 patients with native knee septic arthritis undergoing I&D were identified, 8,063 arthroscopic (56.1%) and 6,302 open (43.9%). The mean follow-up was 148 days (interquartile range 53 to 259). A total of 2,156 patients (15.0%) underwent revision I&D. On multivariate analysis, arthroscopic I&D was associated with a reduction in hospital costs of $5,674 and length of stay of 1.46 days (P < 0.001 for both). Arthroscopic I&D was associated with lower overall complications (odds ratio [OR] 0.63, P < 0.001), need for blood transfusion (OR 0.58, P < 0.001), and wound complications (OR 0.32, P < 0.001). Revision-free survival after index I&D was 95.3% at 3 days, 91.0% at 10 days, 88.3% at 30 days, 86.0% at 90 days, and 84.5% at 180 days. No statistically significant difference was observed between surgical approaches on Cox modeling. DISCUSSION: Risk of revision I&D did not differ between arthroscopic and open I&D; however, arthroscopy was associated with decreased costs, length of stay, and complications. Additional study is necessary to confirm these findings and characterize which patients require an open I&D. LEVEL OF EVIDENCE: III.

National trends in the utilization of lumbar disc replacement for lumbar degenerative disc disease over a 10-year period, 2010 to 2019.

Background: Lumbar fusion (LF) is commonly performed to manage lumbar degenerative disc disease (LDDD) that has failed conservative measures. However, lumbar disc replacement (LDR) procedures are increasingly prevalent and designed to preserve motion in carefully selected patients. Methods: A retrospective cohort study was performed using the National Inpatient Sample (NIS), queried from 2010 to 2019 to identify patients undergoing single and double-level LF or LDR with a diagnosis of LDDD using International Classification of Diseases (ICD) 9th (ICD-9) and 10th (ICD-10) revision diagnostic and procedure codes. Propensity score matching (PSM) with a ratio of 2:1 was performed. All cost estimates reflect reported hospital costs adjusted to December 2019 United States Dollars. Results: A total of 1,129,121 LF cases (99.3%) and 8,049 LDR cases (0.7%) were identified, with 364,637 (32.3%) and 712 (8.8%) comprising two-level surgeries, respectively. 1,712 LDRs were performed in 2010 (1.27% of all), decreasing to 565 in 2013 (0.52%), and increased slightly to 870 in 2019 (0.74%). LDR patients were significantly more likely to be younger (mean age 41.2 vs. 57.1, P<0.001) and healthier (mean ECI 0.88 vs. 1.80, P<0.001). On matched analysis, LDR hospital costs were $4,529 less (P<0.001) and length of stay was 0.65 days shorter (P<0.001) than LF patients. LDR patients had lower rates of any complication (7.0% vs. 13.2%, P<0.001), neurologic complication (3.0% vs. 4.2%, P=0.006), and blood transfusion (3.1% vs. 8.1%, P<0.001) compared to LF patients. Conclusions: The prevalence of LDR procedures decreased from 2010-2017 but began to increase again in 2018 and 2019. Single-level LDR was associated with reduced costs and length of stay (LOS), and lower rates of blood transfusion compared to LF in patients with LDDD.

Lower Risk of Revision Surgery After Arthroscopic Versus Open Irrigation and Débridement for Shoulder Septic Arthritis.

BACKGROUND: In the treatment of native shoulder septic arthritis, the optimal irrigation and débridement modality-arthroscopic versus open-is a matter of controversy. We aim to compare revision-free survival (RFS), complications, and resource utilization between these approaches. METHODS: The National Readmission Database was queried from 2016 to 2019 to identify patients using International Classification of Diseases, 10th revision, diagnostic and procedure codes. Days to revision irrigation and débridement (I&D) were calculated for patients during index admission or subsequent readmissions. Multivariate regression was used for healthcare utilization analysis. Survival analysis was done using Kaplan-Meier analysis and Cox proportional hazard regression. RESULTS: A total of 4,113 patients with native shoulder septic arthritis undergoing I&D were identified, 2,775 arthroscopic (67.5%) and 1,338 open (32.5%). The median follow-up was 170 days (interquartile range 79 to 265). A total of 341 patients (8.3%) underwent revision I&D at a median of 9 days. On multivariate analysis, arthroscopic I&D was associated with a reduction in hospital costs of $4,154 ( P < 0.001) and length of stay of 0.78 days ( P = 0.030). Arthroscopic I&D was associated with reduced blood transfusions (odds ratio 0.69, P = 0.001) and wound complications (odds ratio 0.30, P < 0.001). RFS was 96.4%, 94.9%, 93.3%, and 92.6% for arthroscopic I&D and 94.1%, 92.6%, 90.4%, and 89.0% for open I&D at 10, 30, 90 and 180 days, respectively ( P = 0.00043). On multivariate Cox modeling, arthroscopic I&D was associated with improved survival (hazard ratio 0.67, P = 0.00035). On stratified analysis, arthroscopic I&D was associated with improved RFS in patients aged 65 years or older ( P < 0.001), but RFS was similar in those younger than 65 years ( P = 0.17). CONCLUSION: Risk of revision I&D was markedly lower after arthroscopic I&D compared with open, although the protective benefit was limited to patients aged 65 years or older. Arthroscopy was also associated with decreased costs, length of stay, and complications. Although surgeons must consider specific patient factors, our results suggest that arthroscopic I&D is superior to open I&D. LEVEL OF EVIDENCE: III.

A Risk Calculator for the Prediction of C5 Nerve Root Palsy After Instrumented Cervical Fusion.

BACKGROUND: C5 palsy is a common postoperative complication after cervical fusion and is associated with increased health care costs and diminished quality of life. Accurate prediction of C5 palsy may allow for appropriate preoperative counseling and risk stratification. We primarily aim to develop an algorithm for the prediction of C5 palsy after instrumented cervical fusion and identify novel features for risk prediction. Additionally, we aim to build a risk calculator to provide the risk of C5 palsy. METHODS: We identified adult patients who underwent instrumented cervical fusion at a tertiary care medical center between 2013 and 2020. The primary outcome was postoperative C5 palsy. We developed ensemble machine learning, standard machine learning, and logistic regression models predicting the risk of C5 palsy-assessing discrimination and calibration. Additionally, a web-based risk calculator was built with the best-performing model. RESULTS: A total of 1024 patients were included, with 52 cases of C5 palsy. The ensemble model was well-calibrated and demonstrated excellent discrimination with an area under the receiver-operating characteristic curve of 0.773. The following features were the most important for ensemble model performance: diabetes mellitus, bipolar disorder, C5 or C4 level, surgical approach, preoperative non-motor neurologic symptoms, degenerative disease, number of fused levels, and age. CONCLUSIONS: We report a risk calculator that generates patient-specific C5 palsy risk after instrumented cervical fusion. Individualized risk prediction for patients may facilitate improved preoperative patient counseling and risk stratification as well as potential intraoperative mitigating measures. This tool may also aid in addressing potentially modifiable risk factors such as diabetes and obesity.

Pedicled vastus lateralis myocutaneous flap for sacropelvic defects after wide oncologic resection: Wound complications and outcomes.

BACKGROUND AND OBJECTIVES: Adequate coverage of the soft tissue defects from wide resection of sacropelvic malignancies remains challenging. The vastus lateralis flap has been described for coverage in the setting of trauma and infection. This flap has not been described for coverage of sacropelvic tumor defects. METHODS: This is a retrospective cohort study of adult patients who underwent wide resection of a primary sacropelvic malignancy with reconstruction employing a pedicled vastus lateralis flap at two tertiary care centers. Patient demographics, tumor staging, and rate of complications were assessed. RESULTS: Twenty-eight patients were included, with a median age of 51 years. The most common primary tumor was chondrosarcoma followed by chondroblastic osteosarcoma. The median follow-up was 1.1 years. There were 10 cases of wound infection requiring re-operation and three cases of flap failure. CONCLUSIONS: We describe a pedicled vastus lateralis flap for coverage of defects after wide resection of sacropelvic malignancies. A large proportion of our cohort had independent risk factors for wound complications. Even with a cohort with high baseline risk for wound complications, we show that the use of a pedicled vastus lateralis flap is a safe reconstructive option with a wound complication rate in line with the literature.

Development and external validation of predictive algorithms for six-week mortality in spinal metastasis using 4,304 patients from five institutions.

BACKGROUND CONTEXT: Historically, spine surgeons used expected postoperative survival of 3-months to help select candidates for operative intervention in spinal metastasis. However, this cutoff has been challenged by the development of minimally invasive techniques, novel biologics, and advanced radiotherapy. Recent studies have suggested that a life expectancy of 6 weeks may be enough to achieve significant improvements in postoperative health-related quality of life. PURPOSE: The purpose of this study was to develop a model capable of predicting 6-week mortality in patients with spinal metastases treated with radiation or surgery. STUDY DESIGN/SETTING: A retrospective review was conducted at five large tertiary centers in the United States and Taiwan. PATIENT SAMPLE: The development cohort consisted of 3,001 patients undergoing radiotherapy and/or surgery for spinal metastases from one institution. The validation institutional cohort consisted of 1,303 patients from four independent, external institutions. OUTCOME MEASURES: The primary outcome was 6-week mortality. METHODS: Five models were considered to predict 6-week mortality, and the model with the best performance across discrimination, calibration, decision-curve analysis, and overall performance was integrated into an open access web-based application. RESULTS: The most important variables for prediction of 6-week mortality were albumin, primary tumor histology, absolute lymphocyte, three or more spine metastasis, and ECOG score. The elastic-net penalized logistic model was chosen as the best performing model with AUC 0.84 on evaluation in the independent testing set. On external validation in the 1,303 patients from the four independent institutions, the model retained good discriminative ability with an area under the curve of 0.81. The model is available here: https://sorg-apps.shinyapps.io/spinemetssurvival/. CONCLUSIONS: While this study does not advocate for the use of a 6-week life expectancy as criteria for considering operative management, the algorithm developed and externally validated in this study may be helpful for preoperative planning, multidisciplinary management, and shared decision-making in spinal metastasis patients with shorter life expectancy.

Development of a Machine Learning Algorithm for Prediction of Complications and Unplanned Readmission Following Primary Anatomic Total Shoulder Replacements.

Background: The demand and incidence of anatomic total shoulder arthroplasty (aTSA) procedures is projected to increase substantially over the next decade. There is a paucity of accurate risk prediction models which would be of great utility in minimizing morbidity and costs associated with major post-operative complications. Machine learning is a powerful predictive modeling tool and has become increasingly popular, especially in orthopedics. We aimed to build a ML model for prediction of major complications and readmission following primary aTSA. Methods: A large California administrative database was retrospectively reviewed for all adults undergoing primary aTSA between 2015 to 2017. The primary outcome was any major complication or readmission following aTSA. A wide scope of standard ML benchmarks, including Logistic regression (LR), XGBoost, Gradient boosting, AdaBoost and Random Forest were employed to determine their power to predict outcomes. Additionally, important patient features to the prediction models were indentified. Results: There were a total of 10,302 aTSAs with 598 (5.8%) having at least one major post-operative complication or readmission. XGBoost had the highest discriminative power (area under receiver operating curve AUROC of 0.689) of the 5 ML benchmarks with an area under precision recall curve AURPC of 0.207. History of implant complication, severe chronic kidney disease, teaching hospital status, coronary artery disease and male sex were the most important features for the performance of XGBoost. In addition, XGBoost identified teaching hospital status and male sex as markedly more important predictors of outcomes compared to LR models. Conclusion: We report a well calibrated XGBoost ML algorithm for predicting major complications and 30-day readmission following aTSA. History of prior implant complication was the most important patient feature for XGBoost performance, a novel patient feature that surgeons should consider when counseling patients.

Predictors of reoperation after surgery for spinal epidural abscess.

BACKGROUND CONTEXT: Spinal epidural abscess is a rare but severe condition with high rates of postoperative adverse events. PURPOSE: The objective of the study was to identify independent prognostic factors for reoperation using two datasets: an institutional and national database. STUDY DESIGN/SETTING: Retrospective Review. PATIENT SAMPLE: Database 1: Review of five medical centers from 1993 to 2016. Database 2: The National Surgical Quality Improvement Program (NSQIP) was queried between 2012 and 2016. OUTCOME MEASURES: Thirty-day and ninety-day reoperation rate. METHODS: Two independent datasets were reviewed to identify patients with spinal epidural abscesses undergoing spinal surgery. Multivariate analyses were used to determine independent prognostic factors for reoperation while including factors identified in bivariate analyses. RESULTS: Overall, 642 patients underwent surgery for a spinal epidural abscess in the institutional cohort, with a 90-day unplanned reoperation rate of 19.9%. In the NSQIP database, 951 patients were identified with a 30-day unplanned reoperation rate of 12.3%. On multivariate analysis in the NSQIP database, cervical spine abscess was the only factor that reached significance for 30-day reoperation (OR=1.71, 95% CI=1.11-2.63, p=.02, Area under the curve (AUC)=0.61). On multivariate analysis in the institutional cohort, independent prognostic factors for 30-day reoperation were: preoperative urinary incontinence, ventral location of abscess relative to thecal sac, cervical abscess, preoperative wound infection, and leukocytosis (AUC=0.65). Ninety-day reoperation rate also found hypoalbuminemia as a significant predictor (AUC=0.66). CONCLUSION: Six novel independent prognostic factors were identified for 90-day reoperation after surgery for a spinal epidural abscess. The multivariable analysis fairly predicts reoperation, indicating that there may be additional factors that need to be uncovered in future studies. The risk factors delineated in this study through the use of two large cohorts of spinal epidural abscess patients can be used to improve preoperative risk stratification and patient management.

Machine learning-driven identification of novel patient factors for prediction of major complications after posterior cervical spinal fusion.

PURPOSE: Posterior cervical fusion is associated with increased rates of complications and readmission when compared to anterior fusion. Machine learning (ML) models for risk stratification of patients undergoing posterior cervical fusion remain limited. We aim to develop a novel ensemble ML algorithm for prediction of major perioperative complications and readmission after posterior cervical fusion and identify factors important to model performance. METHODS: This is a retrospective cohort study of adults who underwent posterior cervical fusion at non-federal California hospitals between 2015 and 2017. The primary outcome was readmission or major complication. We developed an ensemble model predicting complication risk using an automated ML framework. We compared performance with standard ML models and logistic regression (LR), ranking contribution of included variables to model performance. RESULTS: Of the included 6822 patients, 18.8% suffered a major complication or readmission. The ensemble model demonstrated slightly superior predictive performance compared to LR and standard ML models. The most important features to performance include sex, malignancy, pneumonia, stroke, and teaching hospital status. Seven of the ten most important features for the ensemble model were markedly less important for LR. CONCLUSION: We report an ensemble ML model for prediction of major complications and readmission after posterior cervical fusion with a modest risk prediction advantage compared to LR and benchmark ML models. Notably, the features most important to the ensemble are markedly different from those for LR, suggesting that advanced ML methods may identify novel prognostic factors for adverse outcomes after posterior cervical fusion.

Acute Colonic Pseudo-Obstruction (Ogilvie Syndrome) After Primary Spinal Fusion: An Analysis of Outcomes and Risk Factors from 2005 to 2014.

BACKGROUND: Ogilvie syndrome (OS) is a rare but serious condition seen in the postoperative period. This was an epidemiologic study using data from the National Inpatient Sample from 2005 to 2014 to look at incidence, risk factors, and outcomes associated with OS after primary spine fusion. METHODS: International Classification of Diseases, Ninth Revision codes were used to identify patients who underwent spine fusion surgery. Patients were separated into 2 cohorts based on the diagnosis of OS. Outcome measures and risk factors for cohorts were analyzed using multivariate logistic regression and compared. RESULTS: Over the 10-year study period, 3,884,395 patients underwent primary spine fusion surgery. Among these, 0.04% developed OS during the index hospitalization. The greatest incidence seen in primary fusion involved the thoracic spine (0.15%). OS was more common after spine fusion for spine deformity (P < 0.001). Patients with OS were more likely to be men (P < 0.001), older (P < 0.0001), and have more comorbidities (P < 0.0001). Patients with OS were more likely to require postoperative blood transfusions (odds ratio [OR], 3.39; 95% confidence interval [CI], 2.51-4.59; P < 0.001) and sustain any complication (OR, 4.20; 95% CI, 3.17-5.57; P < 0.001). Patients with OS had a longer length of stay (15.7 vs. 3.9 days; P < 0.001) and increased average hospitalization cost ($63,037.03 vs. $26,792.19; P < 0.001). The development of OS was associated with fluid electrolyte disorder (OR, 4.06; 95% CI, 2.99-5.51; P < 0.001). CONCLUSIONS: OS is a rare but serious complication of primary spine fusion surgery. Identifying the specific risk factors, symptoms, and potential complications related to OS is critical to aid in decreasing the significant morbidity associated with its development.

A Novel, Potentially Universal Machine Learning Algorithm to Predict Complications in Total Knee Arthroplasty.

BACKGROUND: There remains a lack of accurate and validated outcome-prediction models in total knee arthroplasty (TKA). While machine learning (ML) is a powerful predictive tool, determining the proper algorithm to apply across diverse data sets is challenging. AutoPrognosis (AP) is a novel method that uses automated ML framework to incorporate the best performing stages of prognostic modeling into a single well-calibrated algorithm. We aimed to compare various ML methods to AP in predictive performance of complications after TKA. METHODS: Thirty-eight preoperative patient demographics and clinical features from all primary TKAs performed at California-licensed hospitals between 2015 and 2017 were evaluated as predictors of major complications after TKA. Traditional logistic regression (LR), various other ML methods (XGBoost, Gradient Boosting, AdaBoost, and Random Forest), and AP were used for model building to determine discriminative power (area under receiver operating curve), calibration (Brier score), and feature importance. RESULTS: Between 2015 and 2017, there were a total of 156,750 TKAs with 1109 (0.7%) total major complications. AP had the highest discriminative performance with area under receiver operating curve 0.679 compared with LR, XGBoost, Gradient Boosting, AdaBoost, and Random Forest (0.617, 0.601, 0.662, 0.657, and 0.545, respectively). AP (Brier score 0.007) had similar calibration as the other ML methods (0.006, 0.006, 0.022, 0.007, and 0.008, respectively). The variables that are most important for AP differ from those that are most important for LR. CONCLUSION: Compared to conventional ML algorithms, AP has superior discriminative ability with similar calibration and suggests nonlinear relationships between variables in outcomes of TKA.

Prediction of Major Complications and Readmission After Lumbar Spinal Fusion: A Machine Learning-Driven Approach.

BACKGROUND: Given the significant cost and morbidity of patients undergoing lumbar fusion, accurate preoperative risk-stratification would be of great utility. We aim to develop a machine learning model for prediction of major complications and readmission after lumbar fusion. We also aim to identify the factors most important to performance of each tested model. METHODS: We identified 38,788 adult patients who underwent lumbar fusion at any California hospital between 2015 and 2017. The primary outcome was major perioperative complication or readmission within 30 days. We build logistic regression and advanced machine learning models: XGBoost, AdaBoost, Gradient Boosting, and Random Forest. Discrimination and calibration were assessed using area under the receiver operating characteristic curve and Brier score, respectively. RESULTS: There were 4470 major complications (11.5%). The XGBoost algorithm demonstrates the highest discrimination of the machine learning models, outperforming regression. The variables most important to XGBoost performance include angina pectoris, metastatic cancer, teaching hospital status, history of concussion, comorbidity burden, and workers' compensation insurance. Teaching hospital status and concussion history were not found to be important for regression. CONCLUSIONS: We report a machine learning algorithm for prediction of major complications and readmission after lumbar fusion that outperforms logistic regression. Notably, the predictors most important for XGBoost differed from those for regression. The superior performance of XGBoost may be due to the ability of advanced machine learning methods to capture relationships between variables that regression is unable to detect. This tool may identify and address potentially modifiable risk factors, helping risk-stratify patients and decrease complication rates.