Development of Machine Learning Algorithms for Prediction of 30-Day Mortality After Surgery for Spinal Metastasis.

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.

P19. Development and validation of machine learning algorithms for predicting adverse events following surgery for metastatic spine tumors: metastatic adverse events scoring system (MAES).

The aging and multimorbid population and health personnel shortages pose a substantial burden on primary health care. While predictive machine learning (ML) algorithms have the potential to address these challenges, concerns include transparency and insufficient reporting of model validation and effectiveness of the implementation in the clinical workflow. To systematically identify predictive ML algorithms implemented in primary care from peer-reviewed literature and US Food and Drug Administration (FDA) and Conformité Européene (CE) registration databases and to ascertain the public availability of evidence, including peer-reviewed literature, gray literature, and technical reports across the artificial intelligence (AI) life cycle. PubMed, Embase, Web of Science, Cochrane Library, Emcare, Academic Search Premier, IEEE Xplore, ACM Digital Library, MathSciNet, AAAI.org (Association for the Advancement of Artificial Intelligence), arXiv, Epistemonikos, PsycINFO, and Google Scholar were searched for studies published between January 2000 and July 2023, with search terms that were related to AI, primary care, and implementation. The search extended to CE-marked or FDA-approved predictive ML algorithms obtained from relevant registration databases. Three reviewers gathered subsequent evidence involving strategies such as product searches, exploration of references, manufacturer website visits, and direct inquiries to authors and product owners. The extent to which the evidence for each predictive ML algorithm aligned with the Dutch AI predictive algorithm (AIPA) guideline requirements was assessed per AI life cycle phase, producing evidence availability scores. The systematic search identified 43 predictive ML algorithms, of which 25 were commercially available and CE-marked or FDA-approved. The predictive ML algorithms spanned multiple clinical domains, but most (27 [63%]) focused on cardiovascular diseases and diabetes. Most (35 [81%]) were published within the past 5 years. The availability of evidence varied across different phases of the predictive ML algorithm life cycle, with evidence being reported the least for phase 1 (preparation) and phase 5 (impact assessment) (19% and 30%, respectively). Twelve (28%) predictive ML algorithms achieved approximately half of their maximum individual evidence availability score. Overall, predictive ML algorithms from peer-reviewed literature showed higher evidence availability compared with those from FDA-approved or CE-marked databases (45% vs 29%). The findings indicate an urgent need to improve the availability of evidence regarding the predictive ML algorithms' quality criteria. Adopting the Dutch AIPA guideline could facilitate transparent and consistent reporting of the quality criteria that could foster trust among end users and facilitating large-scale implementation.

Simple SummaryThe overall incidence of spinal tumors in the United States was estimated to be 0.62 per 100,000 people. Surgical resection of spinal tumors intends to improve functional status, reduce pain, and, in some patients with isolated metastases or primary tumors, increase survival. Machine learning algorithms show great promise for predicting short-term postoperative outcomes in spinal tumor surgery. With this study, we aim to develop machine learning algorithms for predicting short-term postoperative outcomes and implement these models in an open-source web application.Background: Preoperative prediction of short-term postoperative outcomes in spinal tumor patients can lead to more precise patient care plans that reduce the likelihood of negative outcomes. With this study, we aimed to develop machine learning algorithms for predicting short-term postoperative outcomes and implement these models in an open-source web application. Methods: Patients who underwent surgical resection of spinal tumors were identified using the American College of Surgeons, National Surgical Quality Improvement Program. Three outcomes were predicted: prolonged length of stay (LOS), nonhome discharges, and major complications. Four machine learning algorithms were developed and integrated into an open access web application to predict these outcomes. Results: A total of 3073 patients that underwent spinal tumor resection were included in the analysis. The most accurately predicted outcomes in terms of the area under the receiver operating characteristic curve (AUROC) was the prolonged LOS with a mean AUROC of 0.745 The most accurately predicting algorithm in terms of AUROC was random forest, with a mean AUROC of 0.743. An open access web application was developed for getting predictions for individual patients based on their characteristics and this web application can be accessed here: huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-ST. Conclusion: Machine learning approaches carry significant potential for the purpose of predicting postoperative outcomes following spinal tumor resections. Development of predictive models as clinically useful decision-making tools may considerably enhance risk assessment and prognosis as the amount of data in spinal tumor surgery continues to rise.

Artificial intelligence and spine: rise of the machines

Predicting Surgical Risk in Patients With Cirrhosis: From Art to Science

Precision medicine for traumatic cervical spinal cord injuries: accessible and interpretable machine learning models to predict individualized in-hospital outcomes

Laparoscopic adrenalectomy (LA) has become the standard of care for many conditions requiring removal of the adrenal gland. Previous studies on outcomes after LA have had limitations. This report describes the 30-day morbidity and mortality rates after LA and analyzes factors affecting operative time, hospital length of stay (LOS), and postoperative morbidity. Patients undergoing LA in 2007 and 2008 were identified from the American College of Surgeons' National Surgical Quality Improvement Program (NSQIP). Using multivariate analysis of variance (ANOVA) and logistic regression, 52 demographic/comorbidity variables were analyzed to ascertain factors affecting operative time, LOS, and morbidity. The mean age of the 988 patients was 53.5 ± 13.7 years, and 60% of the patients were women. The mean body mass index (BMI) of the patients was 31.8 ± 7.9 kg/m(2). The 30-day morbidity and mortality rates were 6.8% and 0.5%, respectively. The mean and median operative times were 146.7 ± 66.8 min and 134 min, respectively. The mean and median hospital stays were 2.6 ± 3.1 days and 2 days, respectively. Compared with independent status, totally dependent functional status was associated with a 9.5-day increase in LOS (P = 0.0006) and an increased risk for postoperative morbidity (odds ratio [OR], 14.7; 95% confidence interval [CI], 2.4-91.9; P < 0.0001). Peripheral vascular disease (OR, 7.3; 95% CI, 1.7-31.7; P = 0.008) also was associated with increased 30-day morbidity. Neurologic and respiratory comorbidities were associated with increased LOS (P < 0.05). American Society of Anesthesiology (ASA) class 4 patients had a longer operative time than ASA class 1 patients (P = 0.002). The morbidity and mortality rates after LA are low. Dependent functional status and peripheral vascular disease predispose to postoperative morbidity. Dependent status, higher ASA class, and respiratory and neurologic comorbidities are associated with longer operative time and LOS.

Distal femur fractures remain a significant cause of morbidity and mortality for elderly patients. There is a lack of large population studies investigating short-term outcomes after distal femur c in elderly patients. The purpose of this study is to assess the incidence of and risk factors for various short-term outcomes after distal femur open reduction internal fixation (ORIF) in the geriatric population. The American College of Surgeons' National Surgical Quality Improvement Program (NSQIP) database was used to identify all primary distal femur ORIF cases in patients 60+ years old between January 1, 2015 and December 31, 2020 using Current Procedural Terminology (CPT) codes 27511, 27513, and 27514. Demographic, medical, and surgical variables were extracted for all patients. Propensity score matching was used to match cases in the two age groups based on various demographic and medical comorbidity variables. Several 30-day outcome measures were compared between the 60-79-year-old and 80+-year-old groups both before and after matching. Subsequent multivariate logistic regression was used to identify independent risk factors for 30-day outcome measures in the matched cohort. A total of 2913 patients were included in the final cohort: 1711 patients in the 60-79-year-old group and 1202 patients in the 80+-year-old group. Most patients were female (n = 2385; 81.9%). Prior to matching, the older group had a higher incidence of 30-day mortality (1.9% vs. 6.2%), readmission (3.7% vs. 9.7%, p = 0.024), and non-home discharge (74.3% vs. 89.5%, p < 0.001). Additionally, the older group had a higher rate of blood loss requiring transfusion (30.9% vs. 42.3%, p < 0.001) and medical complications (10.4% vs. 16.4%, p < 0.001), including myocardial infarction (0.7% vs. 2.7%, p < 0.001), pneumonia (2.7% vs. 4.6%, p = 0.008), and urinary tract infection (4.1% vs. 6.1%, p = 0.0188). After matching, the older group consistently had a higher incidence of mortality, non-home-discharge, blood loss requiring transfusion, and myocardial infarction. Various independent risk factors were identified for 30-day morbidity and mortality, including American Society of Anesthesiologists (ASA) classification, body mass index (BMI) status, operative duration, and certain medical comorbidities. Geriatric patients undergoing distal femur ORIF are at significant risk for 30-day morbidity and mortality. After matching, octogenarians and older patients specifically are at increased risk for mortality, non-home discharge, and surgical complications compared to patients aged 60-79 years old. Multiple factors, such as BMI status, ASA classification, operative time, and certain medical comorbidities, are independently associated with poor 30-day outcomes.

Machine Learning-Based Prediction of Short-Term Adverse Postoperative Outcomes in Cervical Disc Arthroplasty Patients

Variability in Reoperation Rates at 182 Hospitals: A Potential Target for Quality Improvement

A Machine Learning Algorithm Outperforms Traditional Multiple Regression to Predict Risk of Unplanned Overnight Stay Following Outpatient Medial Patellofemoral Ligament Reconstruction

Background: The objective of this study was to assess whether resident involvement in distal radius fracture open reduction internal fixation (ORIF) affect 30-day postoperative complication, hospital readmission, reoperation and operative time. Methods: A retrospective study was performed using the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database by querying the Current Procedural Terminology (CPT) codes for distal radius fracture ORIF from 1 January 2011 to 31 December 2014. A final cohort of 5,693 adult patients who underwent distal radius fracture ORIF during the study period were included. Baseline patient demographics and comorbidities, intraoperative factors, including operative time and 30-day postoperative outcomes, including complications, readmission and reoperations, were collected. Bivariate statistical analyses were performed to identify variable associated with complication, readmission, reoperation and operative time. The significance level was adjusted using a Bonferroni correction as multiple comparisons were performed. Results: In this study of 5,693 patients who underwent distal radius fracture ORIF, 66 patients had a complication, 85 patients were readmitted and 61 patients underwent reoperation within 30 days of surgery. Resident involvement in the surgery was not associated with 30-day postoperative complication, readmission or reoperation, but was associated with longer operative time. Moreover, 30-day postoperative complication was associated with older age, American Society of Anesthesiologists (ASA) classification, chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), hypertension and bleeding disorder. Thirty-day readmission was associated with older age, ASA classification, diabetes mellitus, COPD, hypertension, bleeding disorder and functional status. Thirty-day reoperation was associated with higher body mass index (BMI). Longer operative time was associated with younger age, male sex and the absence of bleeding disorder. Conclusions: Resident involvement in distal radius fracture ORIF is associated with longer operative time, but no difference in rates of episode-of-care adverse events. Patients may be reassured that resident involvement in distal radius fracture ORIF does not negatively impact short-term outcomes. Level of Evidence: Level IV (Therapeutic).

To determine the risk factors associated with loss of functional independence after obliterative procedures for pelvic organ prolapse (POP). The American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database was used to collect data on women who underwent obliterative vaginal procedures from 2011 to 2016, using current procedural terminology (CPT) codes for LeFort colpocleisis (57120) and vaginectomy (57110). The criterion for loss of functional independence was a transition from a functionally independent status to a dependent status (discharge to a post-care facility) or death within the 30-day postoperative period. Multivariate regression analysis was utilized to determine factors associated with loss of functional independence. A total of 1847 women were included in the analysis. A loss of functional independence was noted in 50 of the 1847 women (2.6%). The women who suffered loss of functional independence were older than those who were independent postoperatively (mean age 79.3years, SD 7.47 vs. 76.7years, SD 8.1, respectively). On multiple logistic regression analysis, age ≥ 80years (OR 2.8, 95% CI 1.4-5.5), American Society of Anesthesiologists (ASA) classification ≥ 3 (OR 2.3, CI 1.1-4.7) and length of stay ≥ 5days (OR 15.2, 95% CI 6.2-37.1) remained significantly associated with an increased risk of loss of functional independence. Age ≥ 80years, ASA classification ≥ 3 and longer length of stay are associated with an increased risk of loss of functional independence after an obliterative procedure for pelvic organ prolapse. Consideration of these factors during the preoperative decision-making process may help improve outcomes in this cohort.

INTRODUCTION: The American Society of Anesthesiologists (ASA) physical status classification system serves as a valuable tool in preoperative risk assessment. However, its role in predicting postoperative outcomes following nerve transfer surgery remains understudied. METHODS: The 2007-2021 from the American College of Surgeons - National Surgical Quality Improvement Program (ACS-NSQIP) dataset was employed to ascertain patients who underwent nerve transfer surgery. ASA classification was dichotomized into two cohorts:ASA&lt;3 and ASA=3. Univariate analysis was conducted to examine the correlation between ASA classification and LOS, as well as postoperative complications. Multivariate logistic regression was employed to evaluate the independent impact of ASA classification on postoperative outcomes while controlling for baseline clinical characteristics. RESULTS: A total of 3209 patients were included in our analysis (43.1% female, mean age 46.5±16.7 years). On t test, ASA class=3 was associated with longer LOS (3.3 vs 1.5 days, p&lt;0.001) compared to ASA&lt;3. Patients with higher ASA class experienced a significantly increased risks of deep surgical site infection, wound dehiscence, unplanned reoperation, pneumonia, unplanned intubation, postoperative ventilator use, progressive renal insufficiency, urinary tract infection, stroke, myocardial infarction, bleeding requiring transfusion. On univariate analysis, ASA=3 was associated with higher risk of overall complications (OR 2.9 CI: 2.5 - 3.4, p&lt;0.001). On multivariate logistic regression controlling for age, BMI, smoking, preoperative transfusion, disseminated cancer, chronic steroid use, modified 5-item frailty index, ASA=3 was an independent predictor of postoperative complications (OR 2.5 CI: 2.0-3.0, p&lt;0.001). CONCLUSIONS: The ASA physical status classification system serves as a prognostic tool predicting hospital LOSs and postoperative outcomes following nerve transfer surgery. Implementation of preoperative risk stratification based on ASA classification may aid in optimizing perioperative care and improving patient outcomes in this surgical population.