Search

Background and purpose — Advancements in software and hardware have enabled the rise of clinical prediction models based on machine learning (ML) in orthopedic surgery. Given their growing popularity and their likely implementation in clinical practice we evaluated which outcomes these new models have focused on and what methodologies are being employed. Material and methods — We performed a systematic search in PubMed, Embase, and Cochrane Library for studies published up to June 18, 2020. Studies reporting on non-ML prediction models or non-orthopedic outcomes were excluded. After screening 7,138 studies, 59 studies reporting on 77 prediction models were included. We extracted data regarding outcome, study design, and reported performance metrics. Results — Of the 77 identified ML prediction models the most commonly reported outcome domain was medical management (17/77). Spinal surgery was the most commonly involved orthopedic subspecialty (28/77). The most frequently employed algorithm was neural networks (42/77). Median size of datasets was 5,507 (IQR 635–26,364). The median area under the curve (AUC) was 0.80 (IQR 0.73–0.86). Calibration was reported for 26 of the models and 14 provided decision-curve analysis. Interpretation — ML prediction models have been developed for a wide variety of topics in orthopedics. Topics regarding medical management were the most commonly studied. Heterogeneity between studies is based on study size, algorithm, and time-point of outcome. Calibration and decision-curve analysis were generally poorly reported.

Background and purpose — External validation of machine learning (ML) prediction models is an essential step before clinical application. We assessed the proportion, performance, and transparent reporting of externally validated ML prediction models in orthopedic surgery, using the Transparent Reporting for Individual Prognosis or Diagnosis (TRIPOD) guidelines. Material and methods — We performed a systematic search using synonyms for every orthopedic specialty, ML, and external validation. The proportion was determined by using 59 ML prediction models with only internal validation in orthopedic surgical outcome published up until June 18, 2020, previously identified by our group. Model performance was evaluated using discrimination, calibration, and decision-curve analysis. The TRIPOD guidelines assessed transparent reporting. Results — We included 18 studies externally validating 10 different ML prediction models of the 59 available ML models after screening 4,682 studies. All external validations identified in this review retained good discrimination. Other key performance measures were provided in only 3 studies, rendering overall performance evaluation difficult. The overall median TRIPOD completeness was 61% (IQR 43–89), with 6 items being reported in less than 4/18 of the studies. Interpretation — Most current predictive ML models are not externally validated. The 18 available external validation studies were characterized by incomplete reporting of performance measures, limiting a transparent examination of model performance. Further prospective studies are needed to validate or refute the myriad of predictive ML models in orthopedics while adhering to existing guidelines. This ensures clinicians can take full advantage of validated and clinically implementable ML decision tools.

BACKGROUND: Machine learning (ML) is a subdomain of artificial intelligence that enables computers to abstract patterns from data without explicit programming. A myriad of impactful ML applications already exists in orthopaedics ranging from predicting infections after surgery to diagnostic imaging. However, no systematic reviews that we know of have compared, in particular, the performance of ML models with that of clinicians in musculoskeletal imaging to provide an up-to-date summary regarding the extent of applying ML to imaging diagnoses. By doing so, this review delves into where current ML developments stand in aiding orthopaedists in assessing musculoskeletal images. QUESTIONS/PURPOSES: This systematic review aimed (1) to compare performance of ML models versus clinicians in detecting, differentiating, or classifying orthopaedic abnormalities on imaging by (A) accuracy, sensitivity, and specificity, (B) input features (for example, plain radiographs, MRI scans, ultrasound), (C) clinician specialties, and (2) to compare the performance of clinician-aided versus unaided ML models. METHODS: A systematic review was performed in PubMed, Embase, and the Cochrane Library for studies published up to October 1, 2019, using synonyms for machine learning and all potential orthopaedic specialties. We included all studies that compared ML models head-to-head against clinicians in the binary detection of abnormalities in musculoskeletal images. After screening 6531 studies, we ultimately included 12 studies. We conducted quality assessment using the Methodological Index for Non-randomized Studies (MINORS) checklist. All 12 studies were of comparable quality, and they all clearly included six of the eight critical appraisal items (study aim, input feature, ground truth, ML versus human comparison, performance metric, and ML model description). This justified summarizing the findings in a quantitative form by calculating the median absolute improvement of the ML models compared with clinicians for the following metrics of performance: accuracy, sensitivity, and specificity. RESULTS: ML models provided, in aggregate, only very slight improvements in diagnostic accuracy and sensitivity compared with clinicians working alone and were on par in specificity (3% (interquartile range [IQR] -2.0% to 7.5%), 0.06% (IQR -0.03 to 0.14), and 0.00 (IQR -0.048 to 0.048), respectively). Inputs used by the ML models were plain radiographs (n = 8), MRI scans (n = 3), and ultrasound examinations (n = 1). Overall, ML models outperformed clinicians more when interpreting plain radiographs than when interpreting MRIs (17 of 34 and 3 of 16 performance comparisons, respectively). Orthopaedists and radiologists performed similarly to ML models, while ML models mostly outperformed other clinicians (outperformance in 7 of 19, 7 of 23, and 6 of 10 performance comparisons, respectively). Two studies evaluated the performance of clinicians aided and unaided by ML models; both demonstrated considerable improvements in ML-aided clinician performance by reporting a 47% decrease of misinterpretation rate (95% confidence interval [CI] 37 to 54; p < 0.001) and a mean increase in specificity of 0.048 (95% CI 0.029 to 0.068; p < 0.001) in detecting abnormalities on musculoskeletal images. CONCLUSIONS: At present, ML models have comparable performance to clinicians in assessing musculoskeletal images. ML models may enhance the performance of clinicians as a technical supplement rather than as a replacement for clinical intelligence. Future ML-related studies should emphasize how ML models can complement clinicians, instead of determining the overall superiority of one versus the other. This can be accomplished by improving transparent reporting, diminishing bias, determining the feasibility of implantation in the clinical setting, and appropriately tempering conclusions. LEVEL OF EVIDENCE: Level III, diagnostic study.

OBJECTIVEReconstruction of the mobile spine following total en bloc spondylectomy (TES) of one or multiple vertebral bodies in patients with malignant spinal tumors is a challenging procedure with high failure rates. A common reason for reconstructive failure is nonunion, which becomes more problematic when using local radiation therapy. Radiotherapy is an integral part of the management of primary malignant osseous tumors in the spine. Vascularized grafts may help prevent nonunion in the radiotherapy setting. The authors have utilized free vascularized fibular grafts (FVFGs) for reconstruction of the spine following TES. The purpose of this article is to describe the surgical technique for vascularized reconstruction of defects after TES. Additionally, the outcomes of consecutive cases treated with this technique are reported.METHODSThirty-nine patients were treated at the authors’ tertiary care institution for malignant tumors in the mobile spine using FVFG following TES between 2010 and 2018. Postoperative union, reoperations, complications, neurological outcome, and survival were reported. The median follow-up duration was 50 months (range 14–109 months).RESULTSThe cohort consisted of 26 males (67%), and the median age was 58 years. Chordoma was the most prevalent tumor (67%), and the lumbar spine was most affected (46%). Complete union was seen in 26 patients (76%), the overall complication rate was 54%, and implant failure was the most common complication, with 13 patients (33%) affected. In 18 patients (46%), one or more reoperations were needed, and the fixation was surgically revised 15 times (42% of reoperations) in 10 patients (26%). A reconstruction below the L1 vertebra had a higher proportion of implant failure (67%; 8 of 12 patients) compared with higher resections (21%; 5 of 24 patients) (p = 0.011). Graft length, number of resected vertebrae, and docking the FVFG on the endplate or cancellous bone was not associated with union or implant failure on univariate analysis.CONCLUSIONSThe FVFG is an effective reconstruction technique, particularly in the cervicothoracic spine. However, high implant failure rates in the lumbar spine have been seen, which occurred even in cases in which the graft completely healed. Methods to increase the weight-bearing capacity of the graft in the lumbar spine should be considered in these reconstructions. Overall, the rates of failure and revision surgery for FVFG compare with previous reports on reconstruction after TES.

BACKGROUND Patients with bone metastases often are unable to complete quality of life (QoL) questionnaires, and cohabitants (such as spouses, domestic partners, offspring older than 18 years, or other people who live with the patient) could be a reliable alternative. However, the extent of reliability in this complicated patient population remains undefined, and the influence of the cohabitant's condition on their assessment of the patient's QoL is unknown. QUESTIONS/PURPOSES (1) Do QoL scores, measured by the 5-level EuroQol-5D (EQ-5D-5L) version and the Patient-reported Outcomes Measurement Information System (PROMIS) version 1.0 in three domains (anxiety, pain interference, and depression), reported by patients differ markedly from scores as assessed by their cohabitants? (2) Do cohabitants' PROMIS-Depression scores correlate with differences in measured QoL results? METHODS This cross-sectional study included patients and cohabitants older than 18 years of age. Patients included those with presence of histologically confirmed bone metastases (including lymphoma and multiple myeloma), and cohabitants must have been present at the clinic visit. Patients were eligible for inclusion in the study regardless of comorbidities, prognosis, prior surgery, or current treatment. Between June 1, 2016 and March 1, 2017 and between October 1, 2017 and February 26, 2018, all 96 eligible patients were approached, of whom 49% (47) met the selection criteria and were willing to participate. The included 47 patient-cohabitant pairs independently completed the EQ-5D-5L and the eight-item PROMIS for three domains (anxiety, pain, and depression) with respect to the patients' symptoms. The cohabitants also completed the four-item PROMIS-Depression survey with respect to their own symptoms. RESULTS There were no clinically important differences between the scores of patients and their cohabitants for all questionnaires, and the agreement between patient and cohabitant scores was moderate to strong (Spearman correlation coefficients ranging from 0.52 to 0.72 on the four questionnaires; all p values < 0.05). However, despite the good agreement in QoL scores, an increased cohabitant's depression score was correlated with an overestimation of the patient's symptom burden for the anxiety and depression domains (weak Spearman correlation coefficient of 0.33 [95% confidence interval 0.08 to 0.58]; p = 0.01 and moderate Spearman correlation coefficient of 0.52 [95% CI 0.29 to 0.74]; p < 0.01, respectively). CONCLUSION The present findings support that cohabitants might be reliable raters of the QoL of patients with bone metastases. However, if a patient's cohabitant has depression, the cohabitant may overestimate a patient's symptoms in emotional domains such as anxiety and depression, warranting further research that includes cohabitants with and without depression to elucidate the effect of depression on the level of agreement. For now, clinicians may want to reconsider using the cohabitant's judgement if depression is suspected. CLINICAL RELEVANCE These findings suggest that a cohabitant's impressions of a patient's quality of life are, in most instances, accurate; this is potentially helpful in situations where the patient cannot weigh in. Future studies should employ longitudinal designs to see how or whether our findings change over time and with disease progression, and how specific interventions-like different chemotherapeutic regimens or surgery-may factor in.

Background A preoperative estimation of survival is critical for deciding on the operative management of metastatic bone disease of the extremities. Several tools have been developed for this purpose, but there is room for improvement. Machine learning is an increasingly popular and flexible method of prediction model building based on a data set. It raises some skepticism, however, because of the complex structure of these models. Questions/purposes The purposes of this study were (1) to develop machine learning algorithms for 90-day and 1-year survival in patients who received surgical treatment for a bone metastasis of the extremity, and (2) to use these algorithms to identify those clinical factors (demographic, treatment related, or surgical) that are most closely associated with survival after surgery in these patients. Methods All 1090 patients who underwent surgical treatment for a long-bone metastasis at two institutions between 1999 and 2017 were included in this retrospective study. The median age of the patients in the cohort was 63 years (interquartile range [IQR] 54 to 72 years), 56% of patients (610 of 1090) were female, and the median BMI was 27 kg/m (IQR 23 to 30 kg/m). The most affected location was the femur (70%), followed by the humerus (22%). The most common primary tumors were breast (24%) and lung (23%). Intramedullary nailing was the most commonly performed type of surgery (58%), followed by endoprosthetic reconstruction (22%), and plate screw fixation (14%). Missing data were imputed using the missForest methods. Features were selected by random forest algorithms, and five different models were developed on the training set (80% of the data): stochastic gradient boosting, random forest, support vector machine, neural network, and penalized logistic regression. These models were chosen as a result of their classification capability in binary datasets. Model performance was assessed on both the training set and the validation set (20% of the data) by discrimination, calibration, and overall performance. Results We found no differences among the five models for discrimination, with an area under the curve ranging from 0.86 to 0.87. All models were well calibrated, with intercepts ranging from -0.03 to 0.08 and slopes ranging from 1.03 to 1.12. Brier scores ranged from 0.13 to 0.14. The stochastic gradient boosting model was chosen to be deployed as freely available web-based application and explanations on both a global and an individual level were provided. For 90-day survival, the three most important factors associated with poorer survivorship were lower albumin level, higher neutrophil-to-lymphocyte ratio, and rapid growth primary tumor. For 1-year survival, the three most important factors associated with poorer survivorship were lower albumin level, rapid growth primary tumor, and lower hemoglobin level. Conclusions Although the final models must be externally validated, the algorithms showed good performance on internal validation. The final models have been incorporated into a freely accessible web application that can be found at https://sorg-apps.shinyapps.io/extremitymetssurvival/. Pending external validation, clinicians may use this tool to predict survival for their individual patients to help in shared treatment decision making. Level of evidence Level III, therapeutic study.

BACKGROUND CONTEXT Preoperative survival estimation in spinal metastatic disease helps determine the appropriateness of invasive management. The SORG ML 90-day and 1-year machine learning algorithms for survival in spinal metastatic disease were previously developed in a single institutional sample but remain to be externally validated. PURPOSE The purpose of this study was to externally validate these algorithms in an independent population from another institution. STUDY DESIGN/SETTING Retrospective study at a large, tertiary care center. PATIENT SAMPLE Patients 18 years or older who underwent surgery between 2003 and 2016. OUTCOME MEASURES Ninety-day and 1-year mortality. METHODS Baseline characteristics of the validation cohort were compared to the developmental cohort for the SORG ML algorithms. Discrimination (c-statistic and receiver operating curve), calibration (calibration slope, intercept, calibration plot, and observed proportions by predicted risk groups), overall performance (Brier score), and decision curve analysis were used to assess the performance of the SORG ML algorithms in the validation cohort. RESULTS Overall, 176 patients underwent surgery for spinal metastatic disease, of which 44 (22.7%) experienced 90-day mortality and 99 (56.2%) experienced 1-year mortality. The validation cohort differed significantly from the developmental cohort on primary tumor histology, metastatic tumor burden, previous systemic therapy, overall comorbidity burden, and preoperative laboratory characteristics. Despite these differences, the SORG ML algorithms generalized well to the validation cohort on discrimination (c-statistic 0.75–0.81 for 90-day mortality and 0.77–0.78 for 1-year mortality), calibration, Brier score, and decision curve analysis. CONCLUSION and RELEVANCE Initial results from external validation of the SORG ML 90-day and 1-year algorithms for survival prediction in spinal metastatic disease suggest potential utility of these digital decision aids in clinical practice. Further studies are needed to validate or refute these algorithms in large patient samples from prospective, international, multi-institutional trials.

Neutrophil to lymphocyte ratio and platelet to lymphocyte ratio have been previously identified as markers for overall survival in oncology but remain heretofore unexplored in spinal epidural abscess (SEA).The purpose of this study was to determine the impact of these routinely collected assessments on 90-day mortality in SEA.Retrospective, case-control study.Patients 18 years or older diagnosed with SEA at 2 academic medical centers and 3 community hospitals.Ninety-day postdischarge and in-hospital mortality.Complete blood count with differential obtained on the day immediately preceding or on the day of admission was used to calculate platelet to lymphocyte and neutrophil to lymphocyte ratios. Multivariate analyses were used to determine if these ratios were independent risk factors for 90-day mortality.For 1,053 SEA patients included in the study, the rate of 90-day mortality was 134 (12.7%). The rate of 90-day mortality with neutrophil to lymphocyte ratio (≥8) was (20.5%) compared to (8.1%) with neutrophil to lymphocyte ratio8. Neutrophil to lymphocyte ratio was positively associated with bacteremia, elevated erythrocyte sedimentation rate, and concurrent systemic infections (endocarditis, meningitis) and negatively associated with duration of symptoms prior to presentation. On multivariate analysis, elevated neutrophil to lymphocyte remained an independent risk factor for 90-day mortality (odds ratio=2.62, 95% confidence interval=1.66-4.17, p.001). Platelet to lymphocyte ratio was not associated with 90-day mortality.Absolute neutrophil to lymphocyte ratio is a routinely collected but overlooked biomarker in patients with spinal epidural abscess that is a novel independent risk factor for 90-day mortality.

Purpose: An excessive amount of total hospitalization is caused by delays due to patients waiting to be placed in a rehabilitation facility or skilled nursing facility (RF/SNF). An accurate preoperative prediction of who would need a RF/SNF place after surgery could reduce costs and allow more efficient organizational planning. We aimed to develop a machine learning algorithm that predicts non-home discharge after elective surgery for lumbar spinal stenosis. Methods: We used the American College of Surgeons National Surgical Quality Improvement Program to select patient that underwent elective surgery for lumbar spinal stenosis between 2009 and 2016. The primary outcome measure for the algorithm was non-home discharge. Four machine learning algorithms were developed to predict non-home discharge. Performance of the algorithms was measured with discrimination, calibration, and an overall performance score. Results: We included 28,600 patients with a median age of 67 (interquartile range 58–74). The non-home discharge rate was 18.2%. Our final model consisted of the following variables: age, sex, body mass index, diabetes, functional status, ASA class, level, fusion, preoperative hematocrit, and preoperative serum creatinine. The neural network was the best model based on discrimination (c-statistic = 0.751), calibration (slope = 0.933; intercept = 0.037), and overall performance (Brier score = 0.131). Conclusions: A machine learning algorithm is able to predict discharge placement after surgery for lumbar spinal stenosis with both good discrimination and calibration. Implementing this type of algorithm in clinical practice could avert risks associated with delayed discharge and lower costs. Graphical abstract: These slides can be retrieved under Electronic Supplementary Material.[Figure not available: see fulltext.].

Introduction Numerous prognostication models have been developed to estimate survival in patients with extremity metastatic bone disease, but few include albumin despite albumin's role in malnutrition and inflammation. The purpose of this study was to examine two independent datasets to determine the value for albumin in prognosticating survival in this population. Materials and methods Extremity metastatic bone disease patients undergoing surgical management were identified from two independent populations. Population 1: Retrospective chart review at two tertiary care centers. Population 2: A large, national, North American multicenter surgical registry with 30-day follow-up. Bivariate and multivariate analyses were used to examine albumin's value for prognostication at 1-, 3-, and 12-month after surgery. Results In Population 1, 1,090 patients were identified with 1-, 3-, and 12-month mortality rates of 95 (8.8%), 305 (28.9%), and 639 (62.0%), respectively. In Population 2, 1,675 patients were identified with one-month postoperative mortality rates of 148 (8.8%). In both populations, hypoalbuminemia was an independent prognostic factor for mortality at 30 days. In the institutional set, hypoalbuminemia was additionally associated with 3- and 12-month mortality. Conclusions Hypoalbuminemia is a marker for mortality in extremity metastatic bone disease. Further consideration of this marker could improve existing prognostication models in this population. Level of evidence III.

Machine learning (ML) studies are becoming increasingly popular in orthopedics but lack a critically appraisal of their adherence to peer-reviewed guidelines. The objective of this review was to (1) evaluate quality and transparent reporting of ML prediction models in orthopedic surgery based on the transparent reporting of multivariable prediction models for individual prognosis or diagnosis (TRIPOD), and (2) assess risk of bias with the Prediction model Risk Of Bias ASsessment Tool. A systematic review was performed to identify all ML prediction studies published in orthopedic surgery through June 18th, 2020. After screening 7138 studies, 59 studies met the study criteria and were included. Two reviewers independently extracted data and discrepancies were resolved by discussion with at least two additional reviewers present. Across all studies, the overall median completeness for the TRIPOD checklist was 53% (interquartile range 47%-60%). The overall risk of bias was low in 44% (n = 26), high in 41% (n = 24), and unclear in 15% (n = 9). High overall risk of bias was driven by incomplete reporting of performance measures, inadequate handling of missing data, and use of small datasets with inadequate outcome numbers. Although the number of ML studies in orthopedic surgery is increasing rapidly, over 40% of the existing models are at high risk of bias. Furthermore, over half incompletely reported their methods and/or performance measures. Until these issues are adequately addressed to give patients and providers trust in ML models, a considerable gap remains between the development of ML prediction models and their implementation in orthopedic practice.

None of the existing prognostic scoring systems for spinal epidural abscess (SEA) include albumin despite albumin's established role in inflammation, nutrition, lipid peroxidation, and regulation of apoptosis. The purpose of the present study was to determine the prognostic value of albumin in SEA.We performed a retrospective, case-control study of 2 independent data sets: patients with SEA in an institutional population and patients in the National Surgical Quality Improvement Program (NSQIP). Bivariate analyses and multivariate analyses were used to determine whether albumin is an independent prognostic factor for survival in both data sets.For the 1053 patients with SEA in the institutional cohort, the 90-day postdischarge mortality was 134 (12.7%). Overall, 633 (60.1%) underwent surgery in the initial admission, with a 30-day postoperative mortality rate of 5.5% (n = 35). For the 1154 patients with SEA in the NSQIP database, the 30-day postoperative mortality rate was 3.6% (n = 42). The rate of 90-day postdischarge mortality in the institutional cohort for patients with albumin2.3 g/dL was 25.1%. In contrast, the rate for patients with albumin3.3 g/dL was 4.5%. On multivariate analysis of the NSQIP database, hypoalbuminemia was an independent prognostic factor for 30-day postoperative mortality. On multivariate analysis of the institutional cohort, hypoalbuminemia remained a prognostic factor for 90-day postdischarge mortality.Albumin was validated as an independent prognostic factor in patients with SEA. The lack of this marker in existing scoring systems underscores the need for updated models to optimize risk stratification and shared decision-making before surgery.

Background Determination of the appropriateness of invasive management in patients with spinal metastatic disease requires accurate pre-operative estimation of survival. The purpose of this study was to examine serum alkaline phosphatase as a prognostic marker in spinal metastatic disease. Methods Chart reviews from two tertiary care centres were used to identify spinal metastatic disease patients. Bivariate and multivariate analyses were used to determine if serum alkaline phosphatase was an independent prognostic marker for survival. Results Overall, 732 patients were included with 90-day and 1-year survival of n = 539 (74.9%) and n = 324 (45.7%), respectively. The 1-year survival of patients in the first quartile of alkaline phosphatase (≤73 IU/L) was 78 (57.8%) compared to 31 (24.0%) for patients in the fourth quartile (>140 IU/L). Preoperative serum alkaline phosphatase levels were significantly elevated in patients with multiple spine metastases, non-spine bone metastasis, and visceral metastasis but not in patients with brain metastasis. On multivariate analysis, elevated serum alkaline phosphatase was identified as an independent prognostic factor for survival in spinal metastatic disease. Conclusion Serum alkaline phosphatase is associated with preoperative metastatic tumour burden and is a biomarker for overall survival in spinal metastatic disease.

Background Preoperative prognostication of short-term postoperative mortality in patients with spinal metastatic disease can improve shared decision making around end-of-life care. Objective To (1) develop machine learning algorithms for prediction of short-term mortality and (2) deploy these models in an open access web application. Methods The American College of Surgeons, National Surgical Quality Improvement Program was used to identify patients that underwent operative intervention for metastatic disease. Four machine learning algorithms were developed, and the algorithm with the best performance across discrimination, calibration, and overall performance was integrated into an open access web application. Results The 30-d mortality for the 1790 patients undergoing surgery for spinal metastatic disease was 8.49%. Preoperative factors used for prognostication were albumin, functional status, white blood cell count, hematocrit, alkaline phosphatase, spinal location (cervical, thoracic, lumbosacral), and severity of comorbid systemic disease (American Society of Anesthesiologist Class). In this population, machine learning algorithms developed to predict 30-d mortality performed well on discrimination (c-statistic), calibration (assessed by calibration slope and intercept), Brier score, and decision analysis. An open access web application was developed for the best performing model and this web application can be found here: https://sorg-apps.shinyapps.io/spinemets/. Conclusion Machine learning algorithms are promising for prediction of postoperative outcomes in spinal oncology and these algorithms can be integrated into clinically useful decision tools. As the volume of data in oncology continues to grow, creation of learning systems and deployment of these systems as accessible tools may significantly enhance prognostication and management.

Background Context: Postoperative morbidity may offset the potential benefits of surgical treatment for spine metastatic disease; hence, risk factors for postoperative complications and reoperations should be taken into considerations during surgical decision-making. In addition, it remains unknown whether complications and reoperations shorten these patients’ survival. Purpose: We aimed to describe and identify factors associated with having a complication within 30 days of index surgery as well as factors associated with having a subsequent reoperation. Furthermore, we assessed the effect of 30-day complications and reoperations on the patients’ postoperative survival, as well as described neurologic changes after surgery. Study Design: Retrospective cohort study. Patient Sample: We included 647 patients 18 years and older who had surgery for metastatic disease in the spine between January 2002 and January 2014 in one of two affiliated tertiary care centers. Outcome Measures: Our primary outcomes were complications within 30 days after surgery and reoperations until final follow-up or death. Methods: We used multivariate logistic regression to identify risk factors for 30-day complications and reoperations. We used the Cox regression analysis to assess the effect of postoperative complications and reoperations on survival. Results: From 647 included patients, 205 (32%) had a complication within 30 days. The following variables were independently associated with 30-day complications: lower albumin levels (odds ratio [OR]: 0.69, 95% confidence interval [CI]=0.49–0.96, p=.021), additional comorbidities (OR=1.42, 95% CI=1.00–2.01, p=.048), pathologic fracture (OR=1.41, 95% CI=0.97–2.05, p=.031), three or more spine levels operated upon (OR=1.64, 95% CI=1.02–2.64, p=.027), and combined surgical approach (OR=2.44, 95% CI=1.06–5.60, p=.036). One hundred and fifteen patients (18%) had at least one reoperation after the initial surgery; prior radiotherapy (OR=1.56, 95% CI=1.07–2.29, p=.021) to the spinal tumor was independently associated with reoperation. 30-day complications were associated with worse survival (hazard ratio [HR]=1.40, 95% CI=1.17–1.68, p

Background and objectives Allograft reconstruction of the humerus after resection is preferred by many because of bone stock restoration and biologic attachment of ligaments and muscles to the allograft, theoretically obtaining superior stability and functionality. Our aim was to assess the prevalence of complications and the incidence and etiology for revision surgery in humeral allograft reconstructions. Methods We included patients 18 years and older who underwent wide resection and allograft reconstruction of the humerus for primary and metastatic lesions at our institution between 1990 and 2013. Our primary outcome measures were complications and revision surgery. We used competing risk regression to assess allograft survival. Results Of the 84 patients we included, 47 patients (51%) underwent allograft reconstructions of the proximal humerus, 30 (36%) intercalary, and seven (8%) of the distal humerus. Fifty-one patients (61%) had at least one complication after surgery. Eighteen patients (21%) underwent revision surgery. The 5-year allograft survival was 71%. Conclusion Although allograft reconstructions of the humerus are a valuable option in the orthopedic oncologist's armamentarium, surgeons should mind the accompanying complication rates. Allograft fractures seem to be the main issue for proximal and distal allografts, often leading to revision surgery. Intercalary allografts are mostly troubled by nonunions.

Background Chordomas are locally invasive slow-growing tumors that are difficult to study because of the rarity of the tumors and the lack of significant volumes of patients with longitudinal follow-up. As such, there are currently no machine learning studies in the chordoma literature. The purpose of this study was to develop machine learning models for survival prediction and deploy them as open access web applications as a proof of concept for machine learning in rare nervous system lesions. Methods The National Cancer Institute's Surveillance, Epidemiology, and End Results program database was used to identify adult patients diagnosed with spinal chordoma between 1995 and 2010. Four machine learning models were used to predict 5-year survival for spinal chordoma and assessed by discrimination, calibration, and overall performance. Results The 5-year overall survival for 265 patients with spinal chordoma was 67.5%. Variables used for prediction were age at diagnosis, tumor size, tumor location, extent of tumor invasion, and extent of surgery. For 5-year survival prediction, the Bayes Point Machine achieved the best performance with a c statistic of 0.80, calibration slope of 1.01, calibration intercept of 0.03, and Brier score of 0.16. This model for 5-year mortality prediction was incorporated into an open access application and can be found online ( https://sorg-apps.shinyapps.io/chordoma/ ). Conclusions This analysis of patients with spinal chordoma demonstrated that machine learning models can be developed for survival prediction in rare pathologies and have the potential to serve as the basis for creation of decision support tools in the future.

Background Previous studies have shown that venous thromboembolism (VTE) is a complication associated with neoplastic disease and major orthopaedic surgery. However, many potential risk factors remain undefined. Questions/purposes (1) What proportion of patients develop symptomatic VTE after surgery for long bone metastases? (2) What factors are associated with the development of symptomatic VTE among patients receiving surgery for long bone metastases? (3) Is there an association between the development of symptomatic VTE and 1-year survival among patients undergoing surgery for long bone metastases? (4) Does chemoprophylaxis increase the risk of wound complications among patients undergoing surgery for long bone metastases? Methods A retrospective study identified 682 patients undergoing surgical treatment of long bone metastases between 2002 and 2013 at the Massachusetts General Hospital and Brigham and Women's Hospital. We included patients 18 years of age or older who had a surgical procedure for impending or pathologic metastatic long bone fracture. We considered the humerus, radius, ulna, femur, tibia, and fibula as long bones; metastatic disease was defined as metastases from solid organs, multiple myeloma, or lymphoma. In general, we used 40 mg enoxaparin daily for lower extremity surgery and 325 mg aspirin daily for lower or upper extremity surgery. The primary outcome was a VTE defined as any symptomatic pulmonary embolism (PE) or symptomatic deep vein thrombosis (DVT; proximal and distal) within 90 days of surgery as determined by chart review. The tertiary outcome was defined as any documented wound complication that might be attributable to chemoprophylaxis within 90 days of surgery. At followup after 90 days and 1 year, respectively, 4% (25 of 682) and 8% (53 of 682) were lost to followup. Statistical analysis was performed using multivariable logistic and Cox regression and Kaplan-Meier. Results Overall, 6% (44 of 682) of patients had symptomatic VTE; 22 patients sustained a DVT, and 22 developed a PE. After controlling for relevant confounding variables, higher preoperative hemoglobin level was independently associated (odds ratio [OR], 0.75; 95% confidence interval [CI], 0.60-0.93; p = 0.011) with decreased symptomatic VTE risk, the presence of symptomatic VTE was associated with a worse 1-year survival rate (VTE: 27% [95% CI, 14%-40%] and non-VTE: 39% [95% CI, 35%-43%]; p = 0.041), and no association was found between wound complications and the use of chemoprophylaxis (OR, 3.29; 95% CI, 0.43-25.17; p = 0.252). Conclusions The risk of symptomatic 90-day VTE is high in patients undergoing surgery for long bone metastases. Further study would be needed to determine the VTE prevention strategy that best balances risks and benefits to address this complication. Level of evidence Level III, therapeutic study.

BackgroundSeveral studies have identified prognostic factors for patients with chondrosarcoma, but there are few studies investigating the accuracy of computationally intensive methods such as machine learning. Machine learning is a type of artificial intelligence that enables computers to l

Background Spinal epidural abscess is a high-risk condition that can lead to paralysis or death. It would be of clinical and prognostic utility to identify which subset of patients with spinal epidural abscess is likely to develop a motor deficit or die within 90 days of discharge. Methods We identified all patients ≥18 years of age who were admitted to our hospital system with a diagnosis of spinal epidural abscess during the period of 1993 to 2016. Explanatory variables were collected retrospectively. Bivariate and multivariable logistic regression was performed using these variables to identify independent predictors of motor deficit and 90-day mortality. Nomograms were then constructed to quantify the risk of these outcomes. Results Of the 1,053 patients we identified with spinal epidural abscess, 362 presented with motor weakness. One hundred and thirty-four patients died within 90 days of discharge, inclusive of those who died during hospitalization. Multivariable logistic regression yielded 8 independent predictors of pre-treatment motor deficit and 8 independent predictors of 90-day mortality. We constructed nomograms that generated a probability of pre-treatment motor deficit or 90-day mortality on the basis of the presence of these factors. Conclusions By quantifying the risk of pre-treatment motor deficit and 90-day mortality, our nomograms may provide useful prognostic information for the treatment team. Timely treatment of neurologically intact patients with a high risk of developing a motor deficit is necessary to avoid residual motor weakness and improve survival. Level of evidence Therapeutic Level IV. See Instructions for Authors for a complete description of Levels of Evidence.

STUDY DESIGN This was a retrospective cohort study from 2 affiliated tertiary care referral centers for spine disease. OBJECTIVE The purpose of this article was to assess the prevalence of incidental (ie, asymptomatic) and symptomatic lumbar synovial facet cysts on magnetic resonance imaging. Secondarily, we assessed whether the prevalence increases with age. In addition, we assessed differences in patient and cyst characteristics between asymptomatic and symptomatic facet cysts. SUMMARY OF BACKGROUND The prevalence of symptomatic and asymptomatic synovial facet cysts in the lumbar spine has been incompletely established, and, although many studies demonstrate an association with degenerative spine disease, no cumulative increase in prevalence of synovial facet cysts with increasing age has been presented. METHODS We included 19,010 consecutive patients who underwent a dedicated lumbar spine magnetic resonance imaging between 2004 and 2015. Our outcome measures were symptomatic and asymptomatic facet cysts. A symptomatic cyst was defined as a cyst with symptoms of radiculopathy on the same side as the cyst. RESULTS The overall synovial facet cyst prevalence was 6.5% [95% confidence interval (CI), 6.1-6.8]; 46% of the facet cysts were incidental and 54% were symptomatic. Increased age was independently associated with a higher likelihood of having a synovial facet cyst [odds ratio (per 10 y), 1.24, 95% CI, 1.20-1.29; P

Prompt diagnosis and treatment are critical in spinal epidural abscess, as delay can lead to paralysis or death. The initial management decision for spinal epidural abscess is not always clear, with the literature showing conflicting results. When considering nonoperative management, it is crucial to avoid failure of treatment, given the neurologic compromise incurred through failure. Unfortunately, data regarding risk factors associated with failure are scarce.All patients admitted to our hospital system with a diagnosis of spinal epidural abscess from 1993 to 2016 were identified. Patients who were ≥18 years of age and were initially managed nonoperatively were included. Explanatory variables and outcomes were collected retrospectively. Bivariate and multivariable analyses were performed on these variables to identify independent predictors of failure of nonoperative treatment. A nomogram was constructed to generate a risk of failure based on these predictors.We identified 367 patients who initially underwent nonoperative management. Of these, 99 patients underwent medical management that failed. Multivariable logistic regression yielded 6 independent predictors of failure: a presenting motor deficit, pathologic or compression fracture in affected levels, active malignancy, diabetes mellitus, sensory changes, and dorsal location of abscess. We constructed a nomogram that generates a probability of failure based on the presence of these factors.By quantifying the risk of failure on the basis of the presence of 6 independent predictors of treatment failure, our nomogram may provide a useful tool for the treatment team when weighing the risks and benefits of initial nonoperative treatment compared with operative management.Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.

Background: Elevated serum alkaline phosphatase has been previously studied as a biomarker for progression of metastatic disease and implicated in adverse skeletal events and worsened survival. The purpose of this study was to determine if serum alkaline phosphatase was a predictor of short-term mortality of patients undergoing surgery for spinal metastatic disease. Methods: The American College of Surgeons National Surgical Quality Improvement Program was queried for patients undergoing spinal surgery for metastatic disease. Bivariate and multivariable analyses was undertaken to determine the relationship between serum alkaline phosphatase and 30-day mortality. Results: For the 1788 patients undergoing operative intervention for spinal metastatic disease between 2009 and 2016 the 30-day mortality was 8.49% (n = 151). In patients who survived beyond 30-days after surgery, n = 1627 (91.5%) the median [interquartile range] serum alkaline phosphatase levels were 126.4 [75–138], whereas in patients who had 30-day mortality, the serum alkaline phosphatase levels were 179.8 [114–187]. The optimal cut-off for alkaline phosphatase was determined to be 113 IU/L. On multivariable analysis, elevated serum alkaline phosphatase levels were associated with 30-day mortality (OR 1.61, 95% CI 1.12–2.32, p = 0.011). Conclusion: Elevated preoperative serum alkaline phosphatase is a marker for 30-day mortality in patients undergoing surgery for spinal metastatic disease. Future retrospective and prospective study designs should incorporate assessment of this serum biomarker to better understand the role for serum alkaline phosphatase in improving prognostication in spinal metastatic disease.

Background: Various surgical treatment techniques have been developed to treat capitellar osteochondritis dissecans; however, the optimal technique remains the subject of ongoing debate. Purpose: To evaluate clinical outcomes after arthroscopic debridement and microfracture for advanced capitellar osteochondritis dissecans. Study Design: Case series; Level of evidence, 4. Methods: Between 2008 and 2015, the authors followed 77 consecutive patients (81 elbows) who underwent arthroscopic debridement and microfracture, and loose body removal if needed, for advanced capitellar osteochondritis dissecans. Seventy-one patients (75 elbows) with a minimum follow-up of 1 year were included. The mean age was 16 years (SD, ±3.3 years; range, 11-26 years) and the mean follow-up length was 3.5 years (SD, ±1.9 years; range, 1-8.2 years). Based on CT and/or MRI results, 71 lesions were classified as unstable and 4 as stable. Clinical elbow outcome (pain, function, and social-psychological effect) was assessed using the Oxford Elbow Score (OES) at final follow-up (OES range, 0-48). Range of motion and return to sports were recorded. Multivariable linear regression analysis was performed to determine predictors of postoperative OES. Results: Intraoperatively, there were 3 grade 1 lesions, 2 grade 2 lesions, 10 grade 3 lesions, 1 grade 4 lesion, and 59 grade 5 lesions. The mean postoperative OES was 40.8 (SD, ±8.0). An open capitellar physis was a predictor of better elbow outcome (5.8-point increase; P = .025), as well as loose body removal/grade 4-5 lesions (6.9-point increase; P = .0020) and shorter duration of preoperative symptoms (1.4-point increase per year; P = .029). Flexion slightly improved from 134° to 139° ( P < .001); extension deficit slightly improved from 8° to 3° ( P < .001). Pronation ( P = .47) and supination did not improve ( P = .065). Thirty-seven patients (55%) returned to their primary sport at the same level, and 5 patients (7%) returned to a lower level. Seventeen patients (25%) did not return to sport due to elbow-related symptoms, and 10 patients (13%) did not return due to non–elbow-related reasons. No complications were recorded. Conclusion: Arthroscopic debridement and microfracture for advanced capitellar osteochondritis dissecans provide good clinical results, especially in patients with open growth plate, loose body removal, and shorter duration of symptoms. However, only 62% of patients in this study returned to sports.

Purpose To determine the rate of donor-site morbidity after osteochondral autologous transplantation (OATS) for capitellar osteochondritis dissecans. Methods A literature search was performed in PubMed/MEDLINE, Embase, and Cochrane Library to identify studies up to November 6, 2016. Criteria for inclusion were OATS for capitellar osteochondritis dissecans, reported outcomes related to donor sites, ≥10 patients, ≥1 year follow-up, and written in English. Donor-site morbidity was defined as persistent symptoms (≥1 year) or cases that required subsequent intervention. Patient and harvest characteristics were described, as well as the rate of donor-site morbidity. A random effects model was used to calculate and compare weighted group proportions. Results Eleven studies including 190 patients were included. In eight studies, grafts were harvested from the femoral condyle, in three studies, from either the 5th or 6th costal-osteochondral junction. The average number of grafts was 2 (1–5); graft diameter ranged from 2.6 to 11 mm. In the knee-to-elbow group, donor-site morbidity was reported in 10 of 128 patients (7.8%), knee pain during activity (7.0%) and locking sensations (0.8%). In the rib-to-elbow group, one of 62 cases (1.6%) was complicated, a pneumothorax. The proportion in the knee-to-elbow group was 0.04 (95% CI 0.0–0.15), and the proportion in the rib-to-elbow group was 0.01 (95% CI 0.00–0.06). There were no significant differences between both harvest techniques (n.s.). Conclusions Donor-site morbidity after OATS for capitellar osteochondritis dissecans was reported in a considerable group of patients. Level of evidence Level IV, systematic review of level IV studies. Electronic supplementary material The online version of this article (doi:10.1007/s00167-017-4516-8) contains supplementary material, which is available to authorized users.

BACKGROUND CONTEXT Reconstruction of the mobile spine following total en bloc spondylectomy (TES) of one or multiple vertebral bodies in patients suffering from malignant spinal tumors is a challenging procedure with high failure rates. One of the most common reasons for reconstructive failure is nonunion, which is an even greater problem in the setting of local radiation. Radiation is an integral part of the management of primary malignant bone tumors in the spine at our institution. Vascularized grafts may help prevent nonunion in the setting of radiation therapy. We have routinely utilized free vascularized fibular grafts (FVFG) for the reconstruction of the mobile spine following TES. PURPOSE The purpose of this study is to describe the surgical technique utilized for vascularized reconstruction of defects following TES. In addition, we report the outcomes of a consecutive series of patients treated with this technique. STUDY DESIGN/SETTING Retrospective review of electronic health records at one tertiary care center and an illustrated explanation of the surgical technique. PATIENT SAMPLE Adult patients, 18 years or older, undergoing total en bloc resection of primary malignant tumors of the mobile spine followed by reconstruction using a free vascularized fibula graft. OUTCOME MEASURES The primary outcome was reconstruction failure of the construct. METHODS Thirty-nine patients treated at our tertiary care institution between 2010 and 2018 for a diagnosis of malignant tumor in the mobile spine with the use of FVFG following TES were reviewed. Postoperative union, complications, neurological outcome, reoperations, and survival were reported. The median time of follow-up was 50 months (range, 14 – 109 months). RESULTS The cohort consisted of 26 males and 13 females, with a median age of 58 years. Chordoma was the tumor that occurred most often (67%) and tumors occurred most often in the lumbar spine (46%). Complete union was seen in 26 patients (76%), the overall complication rate was 56%, and instrumentation failure was the most common complication with 26% of patients affected. In total 12 patients (31%) suffered from – instrumentation loosening or breakage, nonunion, graft fracture, or a combination of these – for which ten patients (26%) required revision surgery. The fixation was revised a total of 15 times (42% of reoperations). In 18 patients (46%) one or more reoperations were needed. The overall 1-, 3-, and 5-year survival rate was 95%, 92%, and 87%, respectively. CONCLUSIONS The FVFG is an adequate reconstruction technique of the mobile spine following TES for malignant tumors. Although we see higher union rates and comparable complication rates compared to other techniques, the rate of instrumentation failure was high. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

Objective Age and comorbidity burden of patients going anterior cervical discectomy and fusion (ACDF) have increased significantly over the past 2 decades, resulting in increased expenditures. Non-home discharge after ACDF contributes to increased direct and indirect costs of postoperative care. The purpose of this study was to identify independent prognostic factors for discharge disposition in patients undergoing ACDF. Methods A retrospective review was conducted at 5 medical centers to identify patients undergoing ACDF for degenerative conditions. The primary outcome was non-home discharge. Additional outcomes considered included discharge to rehabilitation and home discharge with services. Bivariate and multivariable analyses were used to identify independent prognostic factors for non-home discharge. Results Of 2070 patients undergoing ACDF, 114 (5.5%) had non-home discharge and 63 (3.0%) had discharge to inpatient rehabilitation. Factors independently associated with non-home discharge included older age, marital status, Medicare insurance, Medicaid insurance, previous spine surgery, myelopathy, preoperative comorbidities (hemiplegia/paraplegia, congestive heart failure, cerebrovascular accident), anemia, and leukocytosis. C-statistic for the overall model was 0.85. Results were relatively similar for patients younger than the age of 65 years as well as for discharge to inpatient rehabilitation and discharge home with services. Conclusions Numerous sociodemographic and clinical characteristics influence the risk of non-home discharge and discharge to inpatient rehabilitation in patients undergoing ACDF. Policy makers and payers should consider these factors when determining appropriate preoperative adjustment for risk-based reimbursements.

BACKGROUND CONTEXT The severity of the opioid epidemic has increased scrutiny of opioid prescribing practices. Spine surgery is a high-risk episode for sustained postoperative opioid prescription. PURPOSE To develop machine learning algorithms for preoperative prediction of sustained opioid prescription after anterior cervical discectomy and fusion (ACDF). STUDY DESIGN/SETTING Retrospective, case-control study at two academic medical centers and three community hospitals. PATIENT SAMPLE Electronic health records were queried for adult patients undergoing ACDF for degenerative disorders between January 1, 2000 and March 1, 2018. OUTCOME MEASURES Sustained postoperative opioid prescription was defined as uninterrupted filing of prescription opioid extending to at least 90–180 days after surgery. METHODS Five machine learning models were developed to predict postoperative opioid prescription and assessed for overall performance. RESULTS Of 2,737 patients undergoing ACDF, 270 (9.9%) demonstrated sustained opioid prescription. Variables identified for prediction of sustained opioid prescription were male sex, multilevel surgery, myelopathy, tobacco use, insurance status (Medicaid, Medicare), duration of preoperative opioid use, and medications (antidepressants, benzodiazepines, beta-2-agonist, angiotensin-converting enzyme-inhibitors, gabapentin). The stochastic gradient boosting algorithm achieved the best performance with c-statistic=0.81 and good calibration. Global explanations of the model demonstrated that preoperative opioid duration, antidepressant use, tobacco use, and Medicaid insurance were the most important predictors of sustained postoperative opioid prescription. CONCLUSIONS One-tenth of patients undergoing ACDF demonstrated sustained opioid prescription following surgery. Machine learning algorithms could be used to preoperatively stratify risk these patients, possibly enabling early intervention to reduce the potential for long-term opioid use in this population.

BACKGROUND: Cancer and spinal surgery are both considered risk factors for venous thromboembolism (VTE). However, the risk of symptomatic VTE for patients undergoing surgery for spine metastases remains undefined. QUESTIONS/PURPOSES: The purposes of this study were to: (1) identify the proportion of patients who develop symptomatic VTE within 90-days of surgical treatment for spine metastases; (2) identify the factors associated with the development of symptomatic VTE among patients receiving surgery for spine metastases; (3) assess the association between the development of postoperative symptomatic VTE and 1-year survival among patients who underwent surgery for spine metastases; and (4) assess if chemoprophylaxis increases the risk of wound complications among patients who underwent surgery for spine metastases. METHODS: Between 2002 and 2014, 637 patients at two hospitals underwent spine surgery for metastases. We considered eligible for analysis adult patients whose procedures were to treat cervical, thoracic, or lumbar metastases (including lymphoma and multiple myeloma). At followup after 90 days and 1 year, respectively, 21 of 637 patients (3%) and 41 of 637 patients (6%) were lost to followup. In general, we used 40 mg of enoxaparin or 5000 IUs subcutaneous heparin every 12 hours. Patients on preoperative chemoprophylaxis continued their initial medication postoperatively. All chemoprophylaxis was started 48 hours after surgery and continued day to day but was discontinued if a bleeding complication developed. Low-molecular-weight heparin (including enoxaparin and dalteparin, in general dosages of respectively 40 mg and 5000 IUs daily) was the most commonly used chemoprophylaxis in 308 patients (48%). Subcutaneous heparin was injected into 127 patients (20%); aspirin was used for 92 patients (14%); and warfarin was administered in 21 patients (3.3%). No form of chemoprophylaxis was prescribed for 89 patients (14%). The primary outcome variable, VTE, was defined as any symptomatic pulmonary embolism (PE) or symptomatic deep venous thromboembolism (DVT) within 90 days of surgery as determined by chart review. The secondary outcome was defined as any documented wound complication within 90 days of surgery that might be attributable to chemoprophylaxis. Statistical analysis was performed using multivariable logistic and Cox regression and Kaplan-Meier. RESULTS: Overall, 72 of 637 patients (11%) had symptomatic VTE; 38 (6%) developed a PE–eight (1.3%) of which were fatal–and 40 (6%) a DVT. After controlling for relevant confounding variables such as age, the modified Charlson Comorbidity Index, visceral metastases, and chemoprophylaxis, longer duration of surgery was independently associated with an increased risk of symptomatic VTE (odds ratio 1.15 for each additional hour of surgery; 95% confidence interval [CI], 1.04-1.28; p = 0.009). After controlling for relevant confounding variables such as age, the modified Charlson Comorbidity Index, visceral metastases, and primary tumor type, patients with symptomatic VTE had a worse 1-year survival rate (VTE, 38%; 95% CI, 27–49 versus nonVTE, 47%; 95% CI, 42–51; p = 0.044). After controlling for relevant confounding variables, no association was found between wound complications and the use of chemoprophylaxis (odds ratio, 1.34; 95% CI, 0.62–2.90; p = 0.459). The overall proportion of patients who developed a wound complication was 10% (66 of 637), including 1.1% (seven of 637) spinal epidural hematomas. CONCLUSIONS: The risk of both symptomatic PE and fatal PE is high in this patient population, and those with symptomatic VTE were less likely to survive 1-year than those who did not, though this may reflect overall infirmity as much as anything else, because many of these patients did not die from VTE-related complications. Further study, such as randomized controlled trials with consistent postoperative VTE screening comparing different chemoprophylaxis regimens, are needed to identify better VTE prevention strategies. LEVEL OF EVIDENCE: Level III, therapeutic study.

There is an increasing number of clinical prediction models in spine surgery based on machine learning. The purpose of this manuscript is to review 1) the outcomes they have focused on, 2) which models have been externally validated, and 3) how studies report their performance measures. Only 2 of 39 prediction models have been externally validated and complications were the most studied topics. Area under the curve (AUC) was the most frequently used performance measure, while calibration and decision-curve analysis are poorly reported. Future studies should focus on external validation of existing models. Publications on the development of models should include all fundamental performance measures and consider including a decision-curve analysis.

In recent years, the volume and complexity of electronic patient-generated health data has increased exponentially leading to unprecedented opportunities for automated extraction of clinical features from free-text medical note and speech. Natural language processing (NLP), a branch of computer science that focuses on enabling computers to process human language, can process this rich resource of data for clinical and research purposes. In this narrative review, we provide the fundamentals of NLP and discuss the potentials and challenges in the spinal community. This lays a foundation for the reader to better understand and appreciate the role that NLP will play in the evolving field of artificial intelligence.

BACKGROUND CONTEXT Spine surgery has been identified as a risk factor for prolonged postoperative opioid use. Preoperative prediction of opioid use could improve risk stratification, shared decision-making, and patient counseling before surgery. PURPOSE The primary purpose of this study was to develop algorithms for prediction of prolonged opioid prescription after surgery for lumbar disc herniation. STUDY DESIGN/SETTING Retrospective, case-control study at five medical centers. PATIENT SAMPLE Chart review was conducted for patients undergoing surgery for lumbar disc herniation between January 1, 2000 and March 1, 2018. OUTCOME MEASURES The primary outcome of interest was sustained opioid prescription after surgery to at least 90 to 180 days postoperatively. METHODS Five models (elastic-net penalized logistic regression, random forest, stochastic gradient boosting, neural network, and support vector machine) were developed to predict prolonged opioid prescription. Explanations of predictions were provided globally (averaged across all patients) and locally (for individual patients). RESULTS Overall, 5,413 patients were identified, with sustained postoperative opioid prescription of 416 (7.7%) at 90 to 180 days after surgery. The elastic-net penalized logistic regression model had the best discrimination (c-statistic 0.81) and good calibration and overall performance; the three most important predictors were: instrumentation, duration of preoperative opioid prescription, and comorbidity of depression. The final models were incorporated into an open access web application able to provide predictions as well as patient-specific explanations of the results generated by the algorithms. The application can be found here: https://sorg-apps.shinyapps.io/lumbardiscopioid/ CONCLUSION Preoperative prediction of prolonged postoperative opioid prescription can help identify candidates for increased surveillance after surgery. Patient-centered explanations of predictions can enhance both shared decision-making and quality of care.

Background Increasing prevalence of metastatic disease has been accompanied by increasing rates of surgical intervention. Current tools have poor to fair predictive performance for intermediate (90-d) and long-term (1-yr) mortality. Objective To develop predictive algorithms for spinal metastatic disease at these time points and to provide patient-specific explanations of the predictions generated by these algorithms. Methods Retrospective review was conducted at 2 large academic medical centers to identify patients undergoing initial operative management for spinal metastatic disease between January 2000 and December 2016. Five models (penalized logistic regression, random forest, stochastic gradient boosting, neural network, and support vector machine) were developed to predict 90-d and 1-yr mortality. Results Overall, 732 patients were identified with 90-d and 1-yr mortality rates of 181 (25.1%) and 385 (54.3%), respectively. The stochastic gradient boosting algorithm had the best performance for 90-d mortality and 1-yr mortality. On global variable importance assessment, albumin, primary tumor histology, and performance status were the 3 most important predictors of 90-d mortality. The final models were incorporated into an open access web application able to provide predictions as well as patient-specific explanations of the results generated by the algorithms. The application can be found at https://sorg-apps.shinyapps.io/spinemetssurvival/. Conclusion Preoperative estimation of 90-d and 1-yr mortality was achieved with assessment of more flexible modeling techniques such as machine learning. Integration of these models into applications and patient-centered explanations of predictions represent opportunities for incorporation into healthcare systems as decision tools in the future.

Background context En bloc resection and reconstruction (EBR) in patients with spinal malignancy aims to achieve local disease control. This is an invasive procedure with significant alterations of the physiological anatomy and subsequently, the spino-pelvic alignment. Sagittal spinal parameters are useful measurements to objectively identify disproportionate alignment on a radiograph. In the field of spinal deformities, there is increasing evidence for a relationship between sagittal alignment and patient reported outcomes. Purpose To determine sagittal spino-pelvic alignment after EBR in patients with spinal malignancies and the effect of these parameters on surgical and patient reported outcomes. Study Design A retrospective case series. Methods We included 35 patients who underwent EBR for spinal malignancies between 2000 and 2018. Radiographic measurements were performed using semi-automatic software; the parameters included were pelvic incidence (PI), sacral slope, pelvic tilt (PT), global tilt and lumbar lordosis. We calculated PI-based Global Alignment and Proportion (GAP) scores and prospective patient reported outcome scores Patient-Reported Outcome Measurement Information System–Physical Function (PROMIS-PF) were used. Results Twenty-one (60%) patients filled out the PROMIS-PF score at a median of 16 months (Interquartile Range (IQR) 4–108) after surgery with a median score of 39 (IQR 32–42), the median GAP score was 7 (IQR 5–9). Bivariate analysis showed no statistically significant relationship between GAP score and instrumentation failure or need for revision surgery. Multivariable analysis of GAP score and PROMIS-PF score corrected for local disease recurrence showed a statistically significant correlation coefficient of −1.721 (p=.026; 95%CI=−3.216, −0.226). Conclusion In this cohort, all patients had a moderate or severe disproportioned spinal alignment after EBR and reconstruction surgery. The degree of sagittal spino-pelvic misalignment after EBR for spinal malignancies seems to be associated with patient reported health status in terms of PROMIS-PF scores. Further research with a larger patient cohort and standardized imaging and follow-up protocols is necessary in order to accurately use sagittal alignment as a predictive value for instrumentation failure and revision surgery.

OBJECTIVEIf not anticipated and prearranged, hospital stay can be prolonged while the patient awaits placement in a rehabilitation unit or skilled nursing facility following elective spine surgery. Preoperative prediction of the likelihood of postoperative discharge to any setting other than home (i.e., nonroutine discharge) after elective inpatient spine surgery would be helpful in terms of decreasing hospital length of stay. The purpose of this study was to use machine learning algorithms to develop an open-access web application for preoperative prediction of nonroutine discharges in surgery for elective inpatient lumbar degenerative disc disorders.METHODSThe American College of Surgeons National Surgical Quality Improvement Program was queried to identify patients who underwent elective inpatient spine surgery for lumbar disc herniation or lumbar disc degeneration between 2011 and 2016. Four machine learning algorithms were developed to predict nonroutine discharge and the best algorithm was incorporated into an open-access web application.RESULTSThe rate of nonroutine discharge for 26,364 patients who underwent elective inpatient surgery for lumbar degenerative disc disorders was 9.28%. Predictive factors selected by random forest algorithms were age, sex, body mass index, fusion, level, functional status, extent and severity of comorbid disease (American Society of Anesthesiologists classification), diabetes, and preoperative hematocrit level. On evaluation in the testing set (n = 5273), the neural network had a c-statistic of 0.823, calibration slope of 0.935, calibration intercept of 0.026, and Brier score of 0.0713. On decision curve analysis, the algorithm showed greater net benefit for changing management over all threshold probabilities than changing management on the basis of the American Society of Anesthesiologists classification alone or for all patients or for no patients. The model can be found here: https://sorg-apps.shinyapps.io/discdisposition/.CONCLUSIONSMachine learning algorithms show promising results on internal validation for preoperative prediction of nonroutine discharges. If found to be externally valid, widespread use of these algorithms via the open-access web application by healthcare professionals may help preoperative risk stratification of patients undergoing elective surgery for lumbar degenerative disc disorders.

STUDY DESIGN A retrospective study. OBJECTIVE The aim of this study was to examine practice variation in the treatment of lumbar spinal stenosis and identify targets for reducing variation. SUMMARY OF BACKGROUND DATA Lumbar spinal stenosis is a degenerative condition susceptible to practice variation. Reducing variation aims to improve quality, increase safety, and lower costs. Establishing differences in surgeons' practices from a single institution can help identify personalized variation. METHODS We identified adult patients first diagnosed with lumbar spinal stenosis between 2003 and 2015 in three hospitals of the same institution with ICD-9 codes.We extracted number of office visits, imaging procedures, injections, electromyographies (EMGs), and surgery within the first year after diagnosis; physical therapy within the first 3 months after diagnosis. Multivariable logistic regression was used to identify factors associated with surgery. The coefficient of variation (CV) was calculated to compare the variation in practice. RESULTS The 10,858 patients we included had an average of 2.5 visits (±1.9), 1.5 imaging procedures (±2.0), 0.03 EMGs (±0.22), and 0.16 injections (±0.53); 36% had at least one surgical procedure and 32% had physical therapy as part of their care. The CV was smallest for number of visits (19%) and largest for EMG (140%).Male sex [odds ratio (OR): 1.23, P

Purpose: We aimed to develop a machine learning algorithm that can accurately predict discharge placement in patients undergoing elective surgery for degenerative spondylolisthesis. Methods: The National Surgical Quality Improvement Program (NSQIP) database was used to select patients that underwent surgical treatment for degenerative spondylolisthesis between 2009 and 2016. Our primary outcome measure was non-home discharge which was defined as any discharge not to home for which we grouped together all non-home discharge destinations including rehabilitation facility, skilled nursing facility, and unskilled nursing facility. We used Akaike information criterion to select the most appropriate model based on the outcomes of the stepwise backward logistic regression. Four machine learning algorithms were developed to predict discharge placement and were assessed by discrimination, calibration, and overall performance. Results: Nine thousand three hundred and thirty-eight patients were included. Median age was 63 (interquartile range [IQR] 54–71), and 63% (n = 5,887) were female. The non-home discharge rate was 18.6%. Our models included age, sex, diabetes, elective surgery, BMI, procedure, number of levels, ASA class, preoperative white blood cell count, and preoperative creatinine. The Bayes point machine was considered the best model based on discrimination (AUC = 0.753), calibration (slope = 1.111; intercept = − 0.002), and overall model performance (Brier score = 0.132). Conclusion: This study has shown that it is possible to create a predictive machine learning algorithm with both good accuracy and calibration to predict discharge placement. Using our methodology, this type of model can be developed for many other conditions and (elective) treatments. Graphical abstract: These slides can be retrieved under Electronic Supplementary Material.[Figure not available: see fulltext.].

Lumbar spinal stenosis is a common condition in the elderly for which costs vary substantially by region. Comparing differences between surgeons from a single institution, thereby omitting regional variation, could aid in identifying factors associated with higher costs and individual drivers of costs. The use of decision aids (DAs) has been suggested as one of the possible tools for diminishing costs and cost variation.(1) To determine factors associated with higher costs for treatment of spinal stenosis in the first year after diagnosis in a single institution; (2) to find individual drivers of costs for providers with higher costs; and (3) to determine if the use of DAs can decrease costs and cost variability.Retrospective cohort study.A total of 10,858 patients in 18 different practices diagnosed with lumbar spinal stenosis between January 2003 and July 2015 in three associated hospitals of a single institution.Mean cost for a patient per provider in US dollars within 1 year after diagnosis of lumbar spinal stenosis.We collected all diagnostic testing, office visits, injections, surgery, and occupational or physical therapy related to lumbar spinal stenosis within 1 year after initial diagnosis. We used multivariable linear regression to determine independent predictors for costs. Providers were grouped in tiers based on mean total costs per patient to find drivers of costs. To assess the DAs effect on costs and cost variability, we matched DA patients one-to-one with non-DA patients.Male gender (β 0.10, 95% confidence interval [CI] 0.05-0.15, p.001), seeing an additional provider (β 0.77, 95% CI 0.69-0.86, p.001), and having an additional spine diagnosis (β 0.79, 95% CI 0.74-0.84, p.001) were associated with higher costs. Providers in the high cost tier had more office visits (p.001), more imaging procedures (p.001), less occupational or physical therapy (p=.002), and less surgery (p=.001) compared with the middle tier. Eighty-two patients (0.76%) received a DA as part of their care; there was no statistically significant difference between the DA group and the matched group in costs (p=.975).Male gender, seeing an additional provider, and having an additional spine diagnosis were independently associated with higher costs. The main targets for cost reduction we found are imaging procedures and number of office visits. Decision aids were not found to affect cost.

Background Context Recurrence of spinal epidural abscess (SEA) after treatment is an important cause of continued morbidity for patients. Purpose The purpose of this study was to identify independent predictors of recurrence of SEA. Study Design/Setting This was a retrospective, case-control study. Patient Sample Patients 18 years or older with a diagnosis of SEA admitted to our hospital system during the study period were included in the study sample. Outcome Measures The outcome measure was recurrence of SEA, defined as a reaccumulation of pus or infected granulation tissue in the epidural space after initial treatment. Methods All patients older than 18 years admitted to our hospital system with a diagnosis of SEA from 1993 to 2016 were identified, and explanatory variables and outcomes were collected retrospectively. Patients 18 years or older diagnosed with SEA were included. We excluded patients whose treatment was initiated at an outside institution. Bivariate and multivariate analyses were performed to identify independent predictors of recurrence. Results We identified 1,053 patients with SEA. We only considered patients to be recurrence-free if they had no documented recurrence with greater than 20 weeks of follow-up. Five hundred thirty-four patients were recurrence-free and 38 had documented recurrence, yielding 572 patients who were included in this analysis. Bivariate and multivariate analyses identified three independent predictors of recurrence: history of intravenous drug use, fecal incontinence or retention, and local spinal wound infection. Conclusions Patients with SEA who have a history of intravenous drug use, bowel dysfunction at presentation, or concurrent local spinal wound infection are at increased risk of disease recurrence. These patients ought to be closely followed up after discharge, with frequent serial imaging and aggressive antibiotic treatment.

BACKGROUND CONTEXT The prevalence of symptomatic and asymptomatic synovial facet cysts in the lumbar spine has been incompletely established, and although many studies demonstrate an association with degenerative spine disease, no cumulative increase in prevalence of synovial facet cysts with increasing age has been presented. PURPOSE To assess the prevalence of incidental (ie, asymptomatic) and symptomatic lumbar synovial facet cysts on MRI. Secondarily, we will assess if the prevalence of lumbar synovial facet cysts increases with age and what other factors are associated with its prevalence. In addition, we will assess differences in patient and cyst characteristics between asymptomatic and symptomatic facet cysts. STUDY DESIGN/SETTING Retrospective cohort study from two affiliated tertiary care referral centers for spine disease. PATIENT SAMPLE A total of 19,010 consecutive patients who underwent a dedicated lumbar spine MRI between 2004 and 2015. OUTCOME MEASURES Symptomatic and asymptomatic facet cysts. METHODS A symptomatic cyst was defined as a cyst with symptoms of radiculopathy on the same side as the cyst. The prevalence of incidental, symptomatic, and overall synovial facet cysts was reported as a percentage with 95% confidence interval (95% CI). Multivariable logistic regression analysis was used to assess if age, sex, race, and indication for MRI were associated with having a facet cyst. Generalized estimating equation analysis was used to assess differences between symptomatic versus asymptomatic synovial facet cyst: age, sex, race, side, level, cyst size, and cyst location. RESULTS The overall synovial facet cyst prevalence was 6.5% (95% CI: 6.1–6.8) [1,228/19,010]. These 1,228 patients had 1,553 synovial facet cysts of which 46% cysts (721/1,553) were incidental and 54% cysts (832/1,553) were symptomatic. The prevalence of having one or more incidental synovial facet cysts was 2.7% (95% CI: 2.5%–3.0%) [518/19,010], the prevalence of having one or more symptomatic synovial facet cysts was 3.2% (95% CI: 3.0%–3.5%) [617/19,010], and the prevalence of having both one or more incidental and one or more symptomatic synovial facet cysts was 0.49% (95% CI: 0.39%–0.59%) [93/19,010]. Increased age was independently associated with a higher likelihood of having a synovial facet cyst (odds ratio [per 10years increase in age]: 1.24, 95%CI: 1.20–1.29, p CONCLUSIONS Approximately 1 in 15 patients has at least one synovial facet cyst and about half of them are symptomatic and half are asymptomatic. Having a facet cyst — symptomatic and asymptomatic — is strongly associated with increased age supporting the theory that degenerative spine disease underlies development of facet cysts. Large cyst size and anterior location of the cysts are associated with an increased likelihood of having neurological symptoms.

Purpose To identify factors associated with unplanned reoperation of severely injured index fingers and to address the number of amputations after initial repair. Methods In this retrospective study, we included all patients older than 18 years of age who had repair or immediate amputation for combined index finger injury at 2 level I trauma centers and 1 community hospital tied to a level I trauma center between January 2004 and February 2014. Twelve patients were excluded because of inadequate follow-up. Bivariate and multivariable analyses sought factors associated with unplanned reoperation after repair and immediate amputation. Results Among 114 patients with combined injury, 75 were treated with repair and 39 with immediate amputation. A total of 41 patients had an unplanned reoperation, 33 after repair (44%) and 8 after immediate amputation (21%). In multivariable analysis, patients who had a reoperation for fingers other than the index finger were at risk for unplanned reoperation after repair. Women were more likely to have an unplanned reoperation than men, and patients who had a ray amputation were at risk for unplanned reoperation after immediate amputation. Six patients (18%) had amputation after initial repair. Conclusions Surgeons may counsel patients that they are twice as likely to have an unplanned reoperation after a repair for combined injury of the index finger compared with an immediate amputation. Unplanned reoperations were more common among patients with injuries involving multiple fingers. Effective shared decision making is particularly important in this setting given that 1 in 5 repaired index fingers were eventually amputated. Type of study/level of evidence Therapeutic IV.