Risk Modeling Research Articles

Age-period-cohort modeling of oesophageal carcinoma risk in a middle eastern country: 1980-2019.

Understanding of the factors influencing oesophageal cancer trends is crucial. Therefore, this cross-sectional cohort study sought to disentangle the age, period and cohort effects on the trends of oesophageal cancer in Kuwait. The data on incident oesophageal carcinoma cases diagnosed between January 1, 1980, through December 31, 2019, and reference population were obtained. Age-period-cohort (APC) analysis was conducted using a loglinear Poisson regression model. A total of 496 oesophageal carcinoma cases in 12.8 million person-years (i.e. squamous-cell carcinoma, 269, 54.23%), adenocarcinoma,147, 29.64% and unspecified cases, 80,16.13%) were diagnosed. The overall age-standardized incidence rate (per 105 person-years) of oesophageal carcinoma during the study period was 10.51 (95% CI: 6.62-14.41). The APC analysis results showed that the age and birth cohort effects were the significant determinants of declining, and subsequently steadying the oesophageal carcinoma incidence rates. A substantial decline in oesophageal carcinoma incidence rates was recorded, which significantly varied in all three temporal dimensions. The observed birth cohort patterns suggest changing lifestyle and dietary patterns seem to be responsible for decreasing oesophageal carcinoma risk in Kuwait. Future studies may look for the component causes maintaining the endemicity of oesophageal carcinoma risk in this and similar countries in the region.

Computing pth root of a transition matrix with a deep unsupervised learning approach

Transition matrices are a tool used to describe transition probabilities of a Markovian process. From an applied perspective, transition matrices are, for example, used in the fields of credit risk modeling and medical decision analysis. In these domains, it is possible to estimate matrices according to the frequency of data observations, but for modeling purposes, it may be necessary to calculate transition matrices for shorter and more restricted time intervals. From a mathematical perspective, this requirement translates into calculating the pth root of a stochastic matrix through a constrained optimization problem. This article focuses on computing pth root of transition matrices and proposes a novel approach using deep learning-based algorithms, particularly Convolutional Neural Networks. Comparisons with traditional algorithms reveal that, in certain contexts, deep learning models offer computational advantages, especially regarding time efficiency. The applicability of the models is verified through a case study built on Fitch Ratings data.

Analysis of Healthcare Waste Treatment Problem Based on an Interval-Valued Fermatean Fuzzy Decision-Making Method

Medical waste risks people and the environment because it can be infectious and contain hazardous substances. For patients to receive medical attention and recover in a safe environment, garbage must be disposed of correctly. Improper medical waste disposal poses a severe risk to society, which can accelerate the development of various pandemics and epidemics. In this case, medical waste disposal should be handled appropriately. This study presents an integrated multi-criteria decision-making method consisting of entropy, PIPRECIA, and MACROS methods based on an interval-valued Fermatean fuzzy set. This method can guarantee high safety and security for health practitioners and society through effective modeling and ranking of risks associated with medical waste disposal. Five alternatives and eight criteria were determined. According to the results, incineration is the most suitable disposal process for healthcare waste. The performance was then assessed and validated using a sensitivity analysis. In addition, the study's findings and their consequences for lawmakers, businesspeople, technologists, and practitioners are examined. In the future, these stakeholders can concentrate on these deficiencies and provide long-term remedies.

Diversification in Business Valuation

Abstract The importance of the diversification and the related diversification factor in business valuation has not been studied in depth. In CAPM, due to the model assumptions, it is implied that an investor holds a perfectly diversified portfolio, which answers the question of diversification. The objective of this paper is to clarify the definition of diversification and to show how diversification effects can be incorporated into business valuation. Our research contribution entails demonstrating methods to capture the risk situation of an investor based on its asset allocation and then deriving the diversification effect from this assessment. As a methodological approach, we have opted for one that encompasses the most diverse degrees of diversification. Our analysis shows that this approach captures all non-hedged risks and thus performs a stand-alone valuation without diversification effects. If the diversification effect of a specific portfolio is to be included in the valuation, the analyst must shift from a stand-alone approach to a portfolio based approach, where the diversification factor cannot simply be equated with the correlation coefficient but must be calculated using various aggregation methods. Given that the conventional variance-covariance method excludes extreme risks, we propose the use of the copula approach. This approach can enhance risk modeling and yield more accurate results in diversification effect calculations. Our analysis demonstrates that it is possible to effectively model different diversification effects in business valuation. However, this entails substantial data expenses and demands a high degree of methodological expertise from the analyst.

Fragility Modeling of Power Grid Infrastructure for Addressing Climate Change Risks and Adaptation

ABSTRACTThe resilience of electric power grids is threatened by natural hazards. Climate‐related hazards are becoming more frequent and intense due to climate change. Statistical analyses clearly demonstrate a rise in the number of incidents (power failures) and their consequences in recent years. Therefore, it is of utmost importance to understand and quantify the resilience of the infrastructure to external stressors, which is essential for developing efficient climate change adaptation strategies. To accomplish this, robust fragility and other vulnerability models are necessary. These models are employed to assess the level of asset damage and to quantify losses for given hazard intensity measures. In this context, a comprehensive literature review is carried out to shed light on existing fragility models specific to the transmission network, distribution network, and substations. The review is organized into three main sections: damage assessment, fragility curves, and recommendations for climate change adaptation. The first section provides a comprehensive review of past incidents, their causes, and failure modes. The second section reviews analytical and empirical fragility models, emphasizing the need for further research on compound and non‐compound hazards, especially windstorms, floods, lightning, and wildfires. Finally, the third section examines risk mitigation and adaptation strategies in the context of climate change. This review aims to improve the understanding of approaches to enhance the resilience of power grid assets in the face of climate change. These insights are valuable to various stakeholders, including risk analysts and policymakers, who are involved in risk modeling and developing adaptation strategies.

Machine learning-based variable selection for clustered credit risk modeling

AbstractSeveral studies have demonstrated the high prediction accuracy of clustered credit risk modeling. In clustered modeling, borrowers are segmented based on their similarities through cluster analysis, and a separate predictive model is developed for each cluster, resulting in increased predictive accuracy. Unambiguously, its effectiveness depends on the quality of the segmentation, which in turn depends primarily on the choice of variables used in the cluster analysis. However, appropriate variable selection for clustering is a major challenge, particularly for high-dimensional data. In the present study, we propose a machine learning-based variable selection method based on theoretical and regulatory considerations. Formally, the most influential risk drivers from a best-in-class machine learning model are identified using Shapley values and employed as clustering variables. Thus, the information of the explanatory variables crucial for the prediction of the dependent variable is already processed during data segmentation, making each individual predictive model more effective. Through a comparative analysis using two real-world credit default datasets, we show that our proposed approach to clustered modeling leads to the highest prediction accuracy among various clustering models.

Importance Sampling for Cost-Optimized Estimation of Burn Probability Maps in Wildfire Monte Carlo Simulations

Background: Wildfire modelers rely on Monte Carlo simulations of wildland fire to produce burn probability maps. These simulations are computationally expensive. Methods: We study the application of importance sampling to accelerate the estimation of burn probability maps, using L2 distance as the metric of deviation. Results: Assuming a large area of interest, we prove that the optimal proposal distribution reweights the probability of ignitions by the square root of the expected burned area divided by the expected computational cost and then generalize these results to the assets-weighted L2 distance. We also propose a practical approach to searching for a good proposal distribution. Conclusion: These findings contribute quantitative methods for optimizing the precision/computation ratio of wildfire Monte Carlo simulations without biasing the results, offer a principled conceptual framework for justifying and reasoning about other computational shortcuts, and can be readily generalized to a broader spectrum of simulation-based risk modeling.

FIRST-TRIMESTER NUCLEAR MAGNETIC RESONANCE-BASED METABOLOMIC PROFILING INCREASES THE PREDICTION OF GESTATIONAL DIABETES MELLITUS.

To investigate whether nuclear magnetic resonance (NMR)-based metabolomic profiling of maternal blood can be used for first-trimester prediction of gestational diabetes mellitus (GDM). This was a prospective study of 20,000 women attending for routine pregnancy care at 11-13 weeks' gestation. Metabolic profiles were assessed using a high-throughput NMR metabolomics platform. To inform translational applications, we focused on a panel of 34 clinically validated biomarkers for detailed analysis and risk modelling. All biomarkers were used to generate a multivariable logistic regression model to predict GDM. Data were split using random seed into a 70%-30% training and validation set, respectively. Performance of the multivariable models was measured by receiver operating characteristic (ROC) curve analysis and detection rates (DRs) at fixed 10% and 20% false positive rates. Calibration for the combined risk model for all GDM was assessed visually through a figure showing the observed incidence against the predicted risk for GDM. A sensitivity analysis was carried out excluding the 64 women in our cohort who were diagnosed with GDM prior to 20 weeks' gestation. The concentration of several metabolomic biomarkers, including cholesterols, triglycerides, fatty acids, and amino acids, were different in the group that developed GDM, compared to those without GDM. Addition of biomarker profile improved the prediction of GDM provided by maternal demographic characteristics and elements of medical history alone (area under the receiver operating characteristics curve [AUROC] 0.790 and detection rate (DR) of 50%, 95%CI 44.3%-55.7%, at 10% false positive rate (FPR) and 63%, 95%CI 57.4%-68.3%, at 20% FPR, to 0.840, 56%, 95%CI 50.3%-61.6% and 73%, 95%CI 67.7%-77.8%, respectively). The performance of combined testing was better for GDM treated by insulin (ROC AUC 0.905, DR 76%, 95%CI 67.5%-83.2%, at 10% FPR and DR 85%, 95%CI 77.4%-90.9%, at 20%) than GDM treated by diet alone (ROC AUC 0.762, DR 47%, 95%CI 37.7%-56.5%, at 10% FPR and DR 64%, 95%CI 54.5%-72.7%, at 20% FPR). In the calibration plot there was good agreement between the observed incidence of GDM against that predicted from the combined risk model. In the sensitivity analysis excluding the women diagnosed with GDM prior to 20 weeks' gestation, there was a negligible difference in ROC AUC in comparison to the results from the entire cohort combined. Addition of NMR-based metabolomic profiling to risk factors can provide first-trimester prediction of GDM.

Credit Risk Modelling for Small and Medium Enterprises (SMEs) Using Machine Learning

This study focuses on developing a credit risk model specifically designed for Small and Medium Enterprises (SMEs) using advanced machine learning techniques. The purpose is to enhance the precision of credit risk assessments, providing financial institutions with reliable tools to gauge the likelihood of default among SMEs. The methodology includes data collection from traditional and alternative sources, followed by feature engineering and model training using algorithms such as logistic regression, decision trees, and neural networks. Results demonstrate significant improvements in predictive accuracy, enabling financial institutions to make data-driven lending decisions and manage risks more effectively. This model not only contributes to financial stability but also promotes increased access to credit for SMEs.

Enhanced Loan Risk Assessment: A Machine Learning Approach

This project develops a loan eligibility prediction system using machine learning, specifically the Random Forest Classifier. The system addresses the limitations of manual loan approval processes, which are time-consuming and prone to human error. A machine learning model is trained on a dataset of loan applications, incorporating features like gender, marital status, income, loan amount, and credit history. This model automatically predicts the eligibility of loan applicants, providing a faster and more objective assessment than traditional methods. The project also includes a user-friendly graphical user interface (GUI) built using Tkinter, allowing users to input applicant information and receive an immediate loan approval or rejection prediction. The model's performance is evaluated using various metrics to ensure accuracy and reliability. The high accuracy achieved by the k-nearest neighbor algorithm demonstrates the effectiveness of the proposed approach. Keywords: Loan Eligibility Prediction, Loan Risk Assessment, Credit Risk Modeling, Machine Learning, Classification, Random Forest, Ensemble Learning, Hyperparameter Tuning, Model Evaluation

Health risk modeling and risk factors of fluorosis in the fluoride endemic village of Maharashtra: A cross-sectional study.

Fluorosis, a debilitating disease, causes dental mottling and bone deformities among other adverse health effects. The current epidemiological study focused on the prevalence and risk factors of fluorosis in the fluoride endemic village, Dongargaon, Maharashtra, India. The clinical diagnosis of subjects showed 96.49% prevalence of dental fluorosis, and 63.16% prevalence of skeletal fluorosis. Duration of exposure, age of subjects and physiological symptoms like chronic joint pain were found to be significantly correlated with skeletal fluorosis (p < 0.05). Fluoride in drinking water showed seasonal variation with highest level in winter (6.54 mg/L) and lowest level in monsoon (0.54 mg/L). Consequently, the mean urinary fluoride concentration was also highest in winter (9.37 mg/L) and lowest in monsoon (3.88 mg/L). Multilogit regression analysis demonstrated that subjects excreting urinary fluoride above 4 mg/L had 7.25 times higher odds of developing skeletal fluorosis (p < 0.05). Health risk assessment showed a high value of hazard quotient, indicating a high probability of adverse health effects. The estimated daily ingestion of 8.80 mg F and a predicted bone uptake of 0.19 mg/kg/day raised significant health concerns for the studied population. Exposure and bone F uptake modeling suggested that the predicted mean fluoride concentrations of 4018.77 ± 1762.45 mg/kg bone ash in asymptomatic, and 5535.12 ± 1526.30 mg/kg bone ash in symptomatic subjects were significantly different (p < 0.05). The comprehensive risk evaluation in the study clearly brings out the high risk of fluoride exposure and adverse effects in Dongargaon village for the first time. Further, urinary fluoride was shown to have significant association with the risk of skeletal fluorosis. The study warrants implementation of immediate control measures at the study site.

FuzzAGG: A Fuzzing-Driven Attack Graph Generation Framework for Industrial Robot Systems

As industrial robot systems (IRS) are increasingly utilized in smart factories, their information security issues have become particularly critical. Attack graphs, an essential system-level risk modeling technique, traditionally rely on predefined risk attributes and exploitation rules for their generation. However, this approach fails to meet the needs for attack graph generation and analysis in environments with missing risk data. To address this issue, this paper proposes a fuzzing-driven attack graph generation framework, FuzzAGG. This framework aims to provide an efficient and accurate method for generating attack graphs under conditions of incomplete risk data, thereby supporting information security analysis and risk assessment of IRS. In this paper, a risk data model (RDM) is constructed using the Meta Attack Language to achieve a structured description of the risk data of IRS. A fuzzing test case generation algorithm based on the MU-SeqGAN model is proposed, which can generate test cases suitable for the state machines of IRS and map them to specific Risk Data Model Objects (RDMOs). Additionally, a conversion unit is designed to integrate all RDMOs into a risk description file, which is then used by the generation unit to construct a graphical attack graph. In performance tests, FuzzAGG is able to achieve automated construction of IRS attack graphs containing 1000 state nodes in 42min and maintain 88% risk coverage. Taking the IRS of a PCB automated production line the effectiveness of the FuzzAGG framework is validated. The results demonstrate that FuzzAGG can automatically generate and validate an attack graph containing 184 attribute nodes and atomic attack nodes in 8min with high operational efficiency, proving the practicality and reliability of this method in automated attack graph generation.

YOLOv8E: an efficient YOLOv8 method for instance segmentation of individual tree crowns in Wellington City, New Zealand

ABSTRACT Instance segmentation is crucial for analysing individual tree crowns in aerial imagery, which plays an important role in forest management, risk modelling, biodiversity modelling, forest health and human wellbeing studies. Traditional instance segmentation methods struggle in diverse rural landscapes where canopy images feature primarily small and medium tree objects, varying from isolated trees to dense forest stands. This paper introduces YOLOv8E, a new and efficient YOLOv8 method, optimised for precise instance segmentation and species classification of tree crowns. This method includes new schemes for selecting candidate positive samples for each instance and a refined network design tailored for small and medium-sized tree crowns. Adjustments in hyperparameters, particularly within the Task-Aligned Assigner, are also discussed to better suit canopy segmentation tasks. Comprehensive experiments conducted on the datasets for Wellington City, Aotearoa New Zealand, demonstrate that YOLOv8E outperforms a number of recent methods, achieving 36.1 and 32.2 in terms of the Box AP and Mask AP metrics respectively.

Macro-financial linkage, endogenous risk premium and monetary policy: evidence from a semi-structural model estimated for Morocco

PurposeThis paper seeks to investigate the amplitude of macro-financial linkage for an emerging country like Morocco.Design/methodology/approachFor this purpose, it presents a semi-structural new-Keynesian model. In the blocks of the former, a risk premium is charged on the lending rate in addition to the policy rate. To identify the macro-financial linkage, the risk premium is considered endogenous. It is represented as a function of the borrower’s probability of default, which is, in turn, a function of the GDP gap. To identify this two-wave relationship, we estimate an ARDL model between 2009Q1 and 2020Q1. Therefore, we integrate the estimation results into the new-Keynesian semi-structural model.FindingsThe results reveal a significant impact of the financial condition on the path of the business cycle. In fact, demand shocks and nonperforming loan shocks (NPLs) are exacerbated by the presence of macro-financial linkage. Under this condition, the amplitude and persistence of the shocks are amplified and extended.Originality/valueThis paper extends the literature on the interconnection between the real and financial economies by considering the endogeneity of the credit risk premium and modeling its dynamics.

Integrated Health Monitoring System using Fingerprint and Disease Risk Prediction

Recent developments in machine learning and biometric systems have had a big influence on healthcare applications. In order to provide a holistic approach to health monitoring, this research introduces a unique method that combines disease risk prediction with fingerprint-based blood group recognition. Based on user-provided health data, the system uses machine learning algorithms to determine the risk of common illnesses including diabetes, cardiovascular disease, and chronic kidney disease (CKD) and convolutional neural networks (CNNs) to identify blood type from fingerprint photos. Through the integration of fingerprint biometric data with health risk modelling, the suggested approach seeks to increase the precision and usability of health forecasts. The system's effectiveness is demonstrated by the evaluation's excellent classification accuracy for blood groups and trustworthy estimates of illness risk.This work demonstrates the possibility of integrating biometric data with predictive health models and offers a promising step toward customized healthcare solutions

Flood fragility and vulnerability functions for residential buildings in the Province of Leyte, Philippines

AbstractThe Philippines experiences frequent flooding, but, despite expansive tools for risk reduction, there remain gaps in understanding generalised relationships between flood events and damage to residential structures for regions outside the nation's capital. This gap has limited the ability to model flood risk and damage without robust functions to link hazards and housing vulnerability. This research draws on 394 household surveys to empirically derive a suite of flood fragility and vulnerability functions for residential structures in the Province of Leyte for light material, elevated light material and masonry structures. The results showed that masonry construction was more resilient to floods compared to light material counterparts. Elevated light material structures also exhibited lower damages at low inundations but tend to fail abruptly at flood depths greater than 3 m. By empirically deriving flood damage functions, the findings contribute to a more localised approach to quantifying housing vulnerability and risk that can be used for catastrophe and risk modelling, with applications for government agencies, the insurance industry and disaster risk researchers. This research lays the foundation for future flood risk mapping with growing significance under climate change.

Abstract 4136280: Inflammation and Cancer Mortality among CVD Individuals: NHANES-Continuous study

Background: Both cancer and cardiovascular disease (CVD) are the leading causes of death worldwide. While previous research showed a relationship between CVD and cancer incidence, limited evidence is available regarding the relationship between inflammation and cancer mortality in CVD patients. Methods: A prospective cohort study using data from the continuous National Health and Nutrition Examination Survey (NHANES) (1999-2016) merged with Medicare and National Death Index (NDI) mortality data until December 31, 2018. We included CVD individuals with no history of cancer at baseline to investigate the relationship between inflammatory markers, including Neutrophil-to-Lymphocyte ratio (NLR), as the main exposure and cancer mortality as the main outcome. Results: We included 4,089 adult individuals representing 127,809,316 individuals (47% females, 72% non-Hispanic Whites, and 5% Hispanics). The risk factor analysis (Table) using the competing risk modeling showed that while inflammatory markers did not show significant associations with cancer mortality in the overall analysis, upon stratification by race/ethnicity, NH-Blacks showed higher cancer mortality with higher WBC count (aHR 5.61, 95% CI 1.66-18.91, p=0.007) and NLR (aHR 2.97, 95% CI 1.19-7.42, p=0.02), but lower cancer mortality with higher neutrophil (aHR 0.08, 95% CI 0.01-0.60, p=0.02) and lymphocyte (aHR 0.34, 95% CI 0.15-0.80, p=0.02) counts. Notably, our NH-Blacks showed lower neutrophils at baseline than others which could be explained by a phenomenon called “Benign Ethnic Neutropenia -BEN-” that still brings more complexity when interpreting NLR. Conclusions: Our study highlights the close relationship between cardiovascular health, inflammation, and cancer mortality. The findings also suggest that race/ethnicity may play a significant role in the relationship between inflammation and cancer mortality among individuals with CVD. Yet, further research is still needed to underpin the biological pathways influencing this relationship.

Abstract 4123540: Cardiovascular Health and Cancer Incidence and Mortality: NHANES-III and -Continuous study

Background: Cancer and cardiovascular disease (CVD) are the leading causes of death worldwide. Limited evidence is available regarding the relationship between CVD, cardiovascular risk factors (CVRF), and cancer mortality; and how CVRF further modify that relationship. Methods: The first part included 438 and 2100 CVD patients aged 65+ from NHANES-III and Continuous (1999-2016) datasets, respectively. Competing risk models with subdistribution hazards ratio (aHR) were used to identify risk factors for cancer incidence. The second part included 44,591 adult individuals from NHANES-Continuous merged with Medicare and National Death Index mortality data until the end of 2018. We included individuals with no history of cancer at baseline. We used competing risk modeling to test the relationship between CVD at baseline and cancer mortality. We also conducted a comprehensive risk factor analysis. Results: In the first part (Figure), females in NHANES-III had lower cancer risk (aHR 0.39, P=0.001) compared to males. Poor physical activity was associated with increased cancer risk in both datasets (aHR 2.59 in NHANES-III, aHR 1.59 in Continuous). In NHANES-Continuous, age (aHR 1.07, P&lt;0.001) and current smoking (aHR 2.55, P=0.001) showed a significant association with developing cancer. No other factors investigated showed significant associations. In the second analysis (Table), the included sample size represented 1,738,423,317 individuals (52% females, 67% non-Hispanic Whites, and 9% Hispanics). Competing risk modeling showed a significantly higher risk of cancer mortality in individuals with CVD (aHR 1.37, 95% CI 1.07-1.76, P=0.01) after adjusting for age, sex, and race/ethnicity. Notably, cancer mortality increased with aging (aHR 1.08, 95%CI 1.05-1.11, P&lt;0.0001), current smoking status (aHR 6.78, 95%CI 3.43-13.42, P&lt;0.0001), and obesity (aHR 2.32, 95%CI 1.13-4.79, P=0.02). Finally, a significant interaction (P=0.034) was found where those with CVD and obesity showed higher cancer mortality than those with normal BMI (aHR, 95%CI 1.03-2.91, P=0.04). Conclusions: Our study highlights the close relationship between cardiovascular health and cancer, including among CVD patients. Our findings suggest that physical activity and obesity may play a significant role in cancer incidence and mortality among the general population and CVD patients. These findings emphasize the need for more proactive approach in the management of shared risk factors for CVD and cancer.

Brief communication: From modelling to reality – flood modelling gaps highlighted by a recent severe storm surge event along the German Baltic Sea coast

Abstract. In October 2023, the Baltic Sea coasts of Germany and Denmark experienced a severe storm surge, predominantly impacting the German state of Schleswig-Holstein and parts of southern Denmark. The surge led to extensive flooding in cities like Flensburg and Schleswig, causing the breaching of at least six (regional) dikes and causing over EUR 200 million in damages in Schleswig-Holstein. By chance, the peak water levels of this storm surge aligned well with those of recent hydrodynamic flood modelling studies of the region. This rare coincidence offers crucial insights for our understanding of flooding impacts, flood management, and modelling. By comparing those studies to the real-world example using extensive media reports, we aim to extract key insights and identify gaps to be tackled in order to improve flood risk modelling in the Baltic Sea region and beyond.

SURG-21. DEVELOPMENT AND VALIDATION OF A CLINICAL RISK SCORE FOR POSTOPERATIVE OUTCOME IN NEWLY DIAGNOSED GLIOBLASTOMA: A REPORT OF THE RANORESECT GROUP

Abstract BACKGROUND Following resection or biopsy, patients with newly diagnosed glioblastoma frequently enter clinical trials. Nuanced risk assessment is warranted to reduce imbalances between study arms. Here, we aim to make use of our RANO classification for extent of resection (I) to analyze the interactive effects of residual tumor with clinical and molecular factors on outcome and (II) to define a postoperative risk assessment tool. METHODS The RANO resect group retrospectively compiled an international, seven-center training cohort of patients with newly diagnosed glioblastoma. The combined associations of residual tumor with molecular or clinical factors and survival were analyzed, and recursive partitioning analysis was performed for risk modeling. The resulting model was prognostically verified in a separate external validation cohort. RESULTS Our training cohort compromised 1003 patients with newly diagnosed IDH-wildtype glioblastoma, including 744 patients who underwent adjuvant radiochemotherapy (TMZ/RT→TMZ). Residual tumor, MGMT promotor methylation status, age, and KPS were prognostic of overall survival and forwarded into regression tree analysis. By combining eleven terminal nodes and individually weighting the prognostic factors, an additive scale (range, 0-9 points) integrating these four variables distinguished patients with low (0-2 points), intermediate (3-5 points), and high risk (6-9 points) for poor postoperative outcome (median overall survival: 24 vs. 16 vs. 6 months; p&amp;lt;0.01). The presence of the three risk groups was confirmed in the subgroups of patients treated with or without RT/TMZ→TMZ. The prognostic value of the risk score was verified in a external single-center validation cohort of newly diagnosed glioblastoma (n = 258, p&amp;lt;0.01). Compared to previously postulated models, goodness-of-fit measurements were superior for our score. CONCLUSIONS The novel RANO risk score serves as an easy-to-use, yet highly prognostic tool for postoperative patient stratification prior to further therapy. The score may serve to guide patient management and reduce imbalances between study arms in prospective trials.