Traditional Risk Models Research Articles

Modelling of the Marine Traffic Situation Complexity

With the growth of traffic density, traditional collision risk models cannot thoroughly describe the complex traffic any more. Therefore, in this paper, a marine traffic situation complexity model, which is based on the relationship between ships, is introduced to metric the complexity of vessel traffic. Firstly, traffic density factor and traffic conflict factors are considered to construct the traffic unit complexity model, reflecting the traffic complexity and traffic characteristics; Secondly, by extending the traffic unit complexity model into the own-ship complexity model, the influences of the positions of other ship on the complexity spatial distribution are analysed; Finally, two scenarios are imported to validate the availability of the complexity model. The verification results show that the complexity can respond to different route structures and traffic situations sensitively.

Commercial Mortgage Termination and Pricing: Theory and Evidence from 1970 to 2015

Commercial mortgage delinquency and default rates increased to unprecedented levels during the 2007–2009 financial crisis. Traditional credit risk models and underwriting models failed to adequately predict the increase in nonperforming loans resulting from precipitous price declines in an increasing number of commercial properties. These effects, in turn, spilled over into the commercial mortgage-backed securities (CMBSs) market. In this paper, we provide a comprehensive analysis of many topics relating to commercial mortgages, including (1) the determinants of mortgage termination and loss severity, (2) the pricing of commercial mortgages and CMBSs, and (3) the role of special servicers. Our synthesis of the extant commercial real estate literature can be useful to borrowers, originators, underwriters, investors, regulators, and policymakers, especially in relation to developing improved credit risk models and better lending standards.

Abstract 17202: Comparative Performance of the R 2 Chads 2 , Chads 2 , and Cha 2 Ds 2 -vasc Scores in Atrial Fibrillation

Background . Existing scores to estimate stroke risk in atrial fibrillation (AF) have shown variable performance. We compared the predictive performance of the R 2 CHADS 2 score, which includes a term for renal dysfunction, to that of two existing stroke risk scores, CHADS 2 and CHA 2 DS 2 -VASc. Methods . We used data from the ORBIT-AF study, a national, prospective, outpatient registry of incident and prevalent AF in patients >18 years. We examined the association between impaired renal function (CrCl<60 mL/min) and 2-year risk of stroke/systemic embolism (SE) in separate Cox proportional hazards models with linear terms for CHADS 2 and CHA 2 DS 2 -VASc scores. We compared discrimination of the three scores using c-indices and evaluated calibration of R 2 CHADS 2 by comparing event rates in ORBIT to published rates from an external clinical trial population (ROCKET-AF) and an observational cohort (ATRIA). Results . We included N=9743 patients enrolled at 174 ORBIT-AF sites. The median age was 75 years (IQR 67-82), 89.5% were white, 42.5% were female, and 76.4% were taking oral anticoagulation (OAC). Over a median follow-up of 2 years, N=214 stroke/systemic embolism (SE) events occurred (1.00 per 100 pt.-years). Impaired renal function was present in 35.4% of patients and was associated (HR; 95% CI) with increased stroke/SE risk in unadjusted models (1.65; 1.27-2.14). This association was attenuated in models adjusting for CHADS 2 (1.21; 0.92, 1.60) or CHA 2 DS 2 -VASc (1.05; 0.79, 1.40). Discrimination (c-index; 95% CI) was similar for R 2 CHADS 2 , CHADS 2 and CHA 2 DS 2 -VASc; results were consistent by baseline OAC use (Table). Stroke/SE event rates in ORBIT were lower than those in two external populations for all levels of R 2 CHADS 2 . Conclusion: In a well-treated community patient population, renal dysfunction did not improve discrimination of traditional embolic risk models. Future studies examining the association between renal dysfunction and outcomes by type of OAC are needed.

Premorbid function, comorbidity, and frailty predict outcomes after ruptured abdominal aortic aneurysm repair

Strategies to improve outcomes for patients with ruptured abdominal aortic aneurysm (rAAA) are becoming more evident. The aging population, however, continues to make the decision to intervene often difficult, especially given that traditional risk models do not reflect issues of aging and frailty. This study aimed to integrate measures of function alongside comorbidity- and frailty-specific factors to determine outcome. Patients treated for a rAAA between January 2006 and April 2014 were assessed. Demographics, mortality, and requirement for care after discharge as well as a variety of measures of function (physical, social, and psychological) and comorbidity were recorded. The primary outcome was 1-year mortality. Outcome models were generated using multivariate logistic regression and were compared with models of vascular frailty and AAA-related outcome. Of 184 patients treated, 108 (59%) underwent an open surgical repair. The overall 30-day and 1-year mortality were 21.5% and 31.4%, respectively, with an overall median hospital length of stay of 13 days (interquartile range, 6-27 days). An optimal logistic regression model for 12-month mortality used Katz score, Charlson score, number of admission medicines, visual impairment, hearing impairment, hemoglobin level, and statin use as predictors, achieving an area under the receiver operating characteristic curve of 0.84. This novel rAAA model incorporating function and comorbidity offers good predictive power for mortality. It is quick to calculate and may ultimately become helpful for both counseling and selection of patients and comparative audit at a time when outcome in patients with rAAA increasingly comes under the spotlight.

Genetics and Genetic Testing in Pancreatic Cancer

Genetic testing of germline DNA is used in patients suspected of being at risk of pancreatic ductal adenocarcinoma (PDAC) to better define the individual's risk and to determine the mechanism of risk. A high genetic risk increases the pretest probability that a biomarker of early cancer is a true positive and warrants further investigation. The highest PDAC risk is generally associated with a hereditary predisposition. However, the majority of PDAC results from complex, progressive gene-environment interactions that currently fall outside the traditional risk models. Over many years, the combination of inflammation, exposure to DNA-damaging toxins, and failed DNA repair promote the accumulation of somatic mutations in pancreatic cells; PDAC risk is further increased by already present oncogenic germline mutations. Predictive models and new technologies are needed to classify patients into more accurate and mechanistic PDAC risk categories that can be linked to improved surveillance and preventative strategies.

Caring for the adult survivor of Hodgkin lymphoma: highlighting the need for care coordination.

Caring for the adult survivor of Hodgkin lymphoma: highlighting the need for care coordination.

The influence of spatial resolution on human health risk co-benefit estimates for global climate policy assessments

Assessment of the ability of climate policies to produce desired improvements in public health through co-benefits of air pollution reduction can consume resources in both time and research funds. These resources increase significantly as the spatial resolution of models increases. In addition, the level of spatial detail available in macroeconomic models at the heart of climate policy assessments is much lower than that available in traditional human health risk modeling. It is therefore important to determine whether increasing spatial resolution considerably affects risk-based decisions; which kinds of decisions might be affected; and under what conditions they will be affected. Human health risk co-benefits from carbon emissions reductions that bring about concurrent reductions in Particulate Matter (PM10) emissions is therefore examined here at four levels of spatial resolution (Uniform Nation, Uniform Region, Uniform County/city, Health Risk Assessment) in a case study of Taiwan as one of the geographic regions of a global macroeceonomic model, with results that are representative of small, industrialized nations within that global model. A metric of human health risk mortality (YOLL, years of life lost in life expectancy) is compared under assessments ranging from a “uniform simulation” in which there is no spatial resolution of changes in ambient air concentration under a policy to a “highly spatially resolved simulation” (called here Health Risk Assessment). PM10 is chosen in this study as the indicator of air pollution for which risks are assessed due to its significance as a co-benefit of carbon emissions reductions within climate mitigation policy. For the policy examined, the four estimates of mortality in the entirety of Taiwan are 747 YOLL, 834 YOLL, 984 YOLL and 916 YOLL, under Uniform Taiwan, Uniform Region, Uniform County and Health Risk Assessment respectively; or differences of 18%, 9%, 7% if the HRA methodology is taken as the baseline. While these differences are small compared to uncertainties in health risk assessment more generally, the ranks of different regions and of emissions categories as the focus of regulatory efforts estimated at these four levels of spatial resolution are quite different. The results suggest that issues of risk equity within a nation might be missed by the lower levels of spatial resolution, suggesting that low resolution models are suited to calculating national cost-benefit ratios but not as suited to assessing co-benefits of climate policies reflecting intersubject variability in risk, or in identifying sub-national regions and emissions sectors on which to focus attention (although even here, the errors introduced by low spatial resolution are generally less than 40%).

Frailty and cardiovascular disease: potential role of gait speed in surgical risk stratification in older adults.

Frailty is a state of late life decline and vulnerability, typified by physical weakness and decreased physiologic reserve. The epidemiology and pathophysiology of frailty share features with those of cardiovascular disease. Gait speed can be used as a measure of frailty and is a powerful predictor of mortality. Advancing age is a potent risk factor for cardiovascular disease and has been associated with an increased risk of adverse outcomes. Older adults comprise approximately half of cardiac surgery patients, and account for nearly 80% of the major complications and deaths following surgery. The ability of traditional risk models to predict mortality and major morbidity in older patients being considered for cardiac surgery may improve if frailty, as measured by gait speed, is included in their assessment. It is possible that in the future frailty assessment may assist in choosing among therapies (e.g., surgical vs. percutaneous aortic valve replacement for patients with aortic stenosis).

STAMP – Holistic system safety approach or just another risk model?

Risk management has a number of accident causation models that have been used for a number of years. Dr. Nancy Leveson (2002) has developed a new model of accidents using a systems approach. The new model is called Systems Theoretic Accident Modeling and Processes (STAMP). It incorporates three basic components: constraints, hierarchical levels of control, and process loops. In this model, accidents are examined in terms of why the controls that were in place did not prevent or detect the hazard(s) and why these controls were not adequate to enforcing the system safety constraints. A STAMP accident analysis is presented and its usefulness in evaluating system safety is compared to more traditional risk models. STAMP is applied to a case study in the oil and gas industry to demonstrate both practicality and validity of the model. The model successfully identified both direct and indirect violations against existing safety constraints that resulted in the accident at each level of the organization.

Abstract 231: Predicting Readmission Risk Following Coronary Revascularization at the Time of Admission

Objective: Reducing readmissions following hospitalization for coronary revascularization is a national priority. Identifying patients at high risk for readmission early in a hospitalization would enable hospitals to target these individuals for enhanced discharge planning. Traditional risk models identify patients based on characteristics that may not be available until after discharge. We sought to compare the model performance for models based on data that is available earlier in the hospitalization. Methods: We developed models to predict 30-day inpatient readmission to our instititution for a cohort of patients who were revascularized at our institution between January 2010 and December 2013. We developed separate models for patients with percutaneous coronary revascularization (PCI) and coronary artery bypass graft surgery (CABG). We developed three models using data available at three different time points in the hospitalization: 1) at admission, 2) early in the hospitalization and 3) at discharge. Candidate variables for the admission model included demographics, comorbidities, and previous utilization within our system. The second model added initial vital signs and laboratory values. The discharge model added discharge vital signs and laboratory values, discharge medications, new comorbidities, and length of stay. We assessed each model using the c-index. Results: Our cohort included 4,941 PCI patients and 1,633 CABG patients. The readmission rate was 8.4% in the PCI group and 14.2% in the CABG group. For both populations, the discriminative ability of the admission model was high (0.805 for PCI, 0.824 for CABG). The addition of laboratory values and vital signs was associated with slight improvement in discrimination (0.812 for PCI, 0.829 for CABG). The addition of data available at the time of discharge further increased model discrimination (0.833 for PCI, 0.848 for CABG). Except for one model, the strongest predictor of readmission was carrying a diagnosis of hypertension at baseline (OR > 2 for all)). Previous hospitalization within six months was a strong predictor among PCI patients (OR 1.18, p<0.001), and having a previous acute myocardial infarction was the strongest predictor among CABG patients (OR 2.46, p = 0.002). Discussion: Risk prediction models based on data available only at discharge minimally improved the performance of models based solely on demographic and utilization data available at the time of admission. These simplified models may be sufficient to identify patients at highest risk of readmission following coronary revascularization early in the hospitlization. This would allow providers and health systems to target high-risk patients with enhanced discharge planning during the course of the hospitalization, and this may improve the ability to avoid readmissions.

Molecular microscope strategy to improve risk stratification in early antibody-mediated kidney allograft rejection.

Antibody-mediated rejection (ABMR) is the leading cause of kidney allograft loss. We investigated whether the addition of gene expression measurements to conventional methods could serve as a molecular microscope to identify kidneys with ABMR that are at high risk for failure. We studied 939 consecutive kidney recipients at Necker Hospital (2004-2010; principal cohort) and 321 kidney recipients at Saint Louis Hospital (2006-2010; validation cohort) and assessed patients with ABMR in the first 1 year post-transplant. In addition to conventional features, we assessed microarray-based gene expression in transplant biopsy specimens using relevant molecular measurements: the ABMR Molecular Score and endothelial donor-specific antibody-selective transcript set. The main outcomes were kidney transplant loss and progression to chronic transplant injury. We identified 74 patients with ABMR in the principal cohort and 54 patients with ABMR in the validation cohort. Conventional features independently associated with failure were donor age and humoral histologic score (g+ptc+v+cg+C4d). Adjusting for conventional features, ABMR Molecular Score (hazard ratio [HR], 2.22; 95% confidence interval [95% CI], 1.37 to 3.58; P=0.001) and endothelial donor-specific antibody-selective transcripts (HR, 3.02; 95% CI, 1.00 to 9.16; P<0.05) independently associated with an increased risk of graft loss. The results were replicated in the independent validation group. Adding a gene expression assessment to a traditional risk model improved the stratification of patients at risk for graft failure (continuous net reclassification improvement, 1.01; 95% CI, 0.57 to 1.46; P<0.001; integrated discrimination improvement, 0.16; P<0.001). Compared with conventional assessment, the addition of gene expression measurement in kidney transplants with ABMR improves stratification of patients at high risk for graft loss.

Integrated Risk Management Process for Building Projects

The objective of this paper is to provide an integrated risk analysis method for building projects which tracks the probabilities of the occurrences of harmful events perceived by the owners from the conceptual phase to the end of the project. The risk management framework is built on traditional risk modeling bases and integrated with financial aspects. This extension of the original concepts defines the project goal by the minimum requirements to be achieved, then objectively identifies risk factors with measurable financial variables. The contribution to the current body of knowledge is threefold. On the one hand, from a project management point of view, the emphasis on the definition of project objectives shifted towards meeting financial goals by integrating relationships of financial minimum expectations at the model level. On the other hand, from a financial aspect, the opportunity of continuous monitoring and reassessment for risk management processes is created. Finally, based on this framework an automatized risk management process can be implemented, which will indicate if adverse processes start in the life of the project and action plans need to be launched. The automated procedure for risk analysis based on the integrated concept delivers a more effective practical tool to reduce project risks, and opens new research areas in construction project management.

Pulmonary embolism risk stratification: where are we heading?

To the Editor : We read with great interest the paper by Sanchez et al. [1], in which the authors suggested that biomarkers and echocardiography findings would provide additional prognostic information to traditional risk models, such as the pulmonary embolism severity index (PESI), in normotensive acute pulmonary embolism (PE) patients. Patients were assigned to low- (PESI I–II), intermediate- (PESI III–IV) and high-risk classes (PESI V) and further stratified according to right ventricle dysfunction markers (right ventricle/left ventricle ratio, and troponin and brain natriuretic peptide (BNP) levels). The authors reported that low-risk PESI patients without any right ventricle dysfunction markers had a significantly lower risk for adverse events than low-risk patients with abnormal right ventricle/left ventricle ratio, troponin and/or BNP levels. However, the same results were not found for the remaining PESI classes (intermediate and high risk). Reliable risk stratification represents the cornerstone of PE management, which should be capable of recognising patients at high risk of death, who could receive specific therapeutic interventions, and those suitable for outpatient management, with important savings in healthcare costs. Although the PESI score has a proven high discriminative performance in the identification of low-risk patients, outperforming other validated PE risk models [2, 3], some authors have recently demonstrated that the 48-h recalculation of PESI can more accurately identify those eligible for outpatient management …

A snapshot of young adults’ noise exposure reveals evidence of ‘Binge Listening’

There have been several previous studies into daily noise exposure levels in modern urban communities, which typically report mean noise exposure levels (LAeq) for adults between 73 and 79dB. In this study, rather than focus on group mean exposures across a wide age range, individual patterns of noise exposure over 4- and 5-day periods were examined in a group of 45 young adults aged 18–35years. The main objective of the study was to determine the extent to which young adults exhibit a ‘binge listening’ pattern of noise exposure, i.e., high weekend leisure noise vs. low weekday work noise exposure. A secondary objective was to identify the types of activities that generate the highest noise exposures. The results showed that although most participants (60%) were exposed to low daily noise levels, 33% of participants exhibited a ‘binge listening’ exposure pattern characterized by one or two high-noise days, usually a Friday, Saturday or Sunday, preceded or followed by much quieter days. The most notable high-noise activities were playing an instrument solo or in a band; attending a nightclub; and attending a pop concert, each of which recorded average noise levels greater than 100dB. Future research is needed to determine whether ‘binge listening’ is more or less harmful than the chronic exposure presupposed in traditional risk models, however, under the equal-energy principle, repeated ‘binge’ noise exposures from weekend visits to nightclubs, live music events and other high-noise events represent a significant risk to long-term hearing health.

Complement-Binding Anti-HLA Antibodies and Kidney-Allograft Survival

BackgroundAnti-HLA antibodies hamper successful transplantation, and activation of the complement cascade is involved in antibody-mediated rejection. We investigated whether the complement-binding capacity of anti-HLA antibodies plays a role in kidney-allograft failure.MethodsWe enrolled patients who received kidney allografts at two transplantation centers in Paris between January 1, 2005, and January 1, 2011, in a population-based study. Patients were screened for the presence of circulating donor-specific anti-HLA antibodies and their complement-binding capacity. Graft injury phenotype and the time to kidney-allograft loss were assessed.ResultsThe primary analysis included 1016 patients. Patients with complement-binding donor-specific anti-HLA antibodies after transplantation had the lowest 5-year rate of graft survival (54%), as compared with patients with non–complement-binding donor-specific anti-HLA antibodies (93%) and patients without donor-specific anti-HLA antibodies (94%) (P<0.001 for both comparisons). The presence of complement-binding donor-specific anti-HLA antibodies after transplantation was associated with a risk of graft loss that was more than quadrupled (hazard ratio, 4.78; 95% confidence interval [CI], 2.69 to 8.49) when adjusted for clinical, functional, histologic, and immunologic factors. These antibodies were also associated with an increased rate of antibody-mediated rejection, a more severe graft injury phenotype with more extensive microvascular inflammation, and increased deposition of complement fraction C4d within graft capillaries. Adding complement-binding donor-specific anti-HLA antibodies to a traditional risk model improved the stratification of patients at risk for graft failure (continuous net reclassification improvement, 0.75; 95% CI, 0.54 to 0.97).ConclusionsAssessment of the complement-binding capacity of donor-specific anti-HLA antibodies appears to be useful in identifying patients at high risk for kidney-allograft loss.

Vascular Protection in People with Diabetes

Vascular Protection in People with Diabetes

Prognostic value of multiple emerging biomarkers in cardiovascular risk prediction in patients with stable cardiovascular disease

BackgroundFew studies have examined simultaneously the prognostic value of traditional and emerging biomarkers including atrial natriuretic peptide (ANP) and brain-type natriuretic peptide (BNP), for major cardiovascular disease (CVD) outcomes in patients with stable CVD, and results are equivocal. Designand Methods: Mid-regional pro-ANP (MR-proANP) and N-Terminal pro-BNP (NT-proBNP), CRP and homocysteine were measured in stable CVD patients (n = 1456; age: 61.8 y) at inclusion in the SU.FOL.OM3 cohort. Prospective association of biomarkers with risk of heart failure, major cardiovascular (non-fatal myocardial infarction, ischemic stroke or death from CVD) or overall cardiovascular event were examined with Cox proportional-hazards analyses. Increase in prediction risk upon addition of biomarker(s) to the traditional risk model was examined by change in C-statistic, NRI and IDI. ResultsDuring follow-up (median: 4.7 y), 40 heart failure, 145 major cardiovascular and 493 overall cardiovascular events were diagnosed. In models adjusted for age, sex, smoking, diabetes, serum creatinine and CVD inclusion criteria, NT-proBNP and CRP associated significantly with heart failure. Both natriuretic peptides predicted the risk of major cardiovascular events in adjusted models; Hazard ratio (HR) and 95% CI for each SD increase in MR-proANP and NT-proBNP were 1.24 (1.04–1.47), and 1.31 (1.09–1.57), respectively. The addition of NT-proBNP to a traditional risk model increased significantly the area-under-curve for heart failure and overall cardiovascular events (by 6 and 12%, respectively); addition of MR-proANP or homocysteine yielded modest (2%) but statistically significant increase for major cardiovascular events. ConclusionNT-proBNP consistently predicted CVD outcomes and may be useful singly or in combination with MR-proANP for risk-stratification in high-risk patients.

Preoperative Cardiac Risk Assessment for Noncardiac Surgery: Defining Costs and Risks

Cardiac risk stratification before noncardiac surgery remains important. Two major areas have been emphaized, namely, cost-effective risk stratification and enhanced identification of high risk populations. Recent studies have highlighted the lack of quality and affordable medical consultation. The indications for resting preoperative echocardiography merit streamlining, given recent data that failed to demonstrate tangible benefit. Further more, noninvasive cardiac stress testing is expensive and unnecessary in low risk patients. Perioperative troponin determination significantly improves the detection of myocardial infarction, facilitating its early management. The revised cardiac risk index is a standard tool for risk stratification, despite multiple limitations. The first approach has been tore calibrate the traditional risk index to specific high-risk surgical subgroups. The second approach has been to develop new cardiac risk models with more power. Both approaches have yielded risk calculators that out perform the traditional risk model. Furthermore, this latest generation of risk models is available as online calculators that can be accessed at the bedside. Further clinical trials are indicated to test the validity, clinical utility, and cost-effectiveness of these novel risk calculators. It is likely that these powerful instruments will refine the indications for specialized cardiac testing, offering multiple opportunities to reduce perioperative risk and cost simultaneously.

Dynamics of credit spread moments of European corporate bond indexes

Traditional quantitative credit risk models assume that changes in credit spreads are normally distributed but empirical evidence shows that they are likely to be skewed, fat-tailed, and change behaviour over time. Not taking into account such characteristics can compromise calculation of loss probabilities, pricing of credit derivatives, and profitability of trading strategies. Therefore, the aim of this study is to investigate the dynamics of higher moments of changes in credit spreads of European corporate bond indexes using extensions of GARCH type models that allow for time-varying volatility, skewness and kurtosis of changes in credit spreads as well as a regime-switching GARCH model which allows for regime shifts in the volatility of changes in credit spreads. Performance evaluation methods are used to assess which model captures the dynamics of observed distribution of the changes in credit spreads, produces superior volatility forecasts and Value-at-Risk estimates, and yields profitable trading strategies. The results presented can have significant implications for risk management, trading activities, and pricing of credit derivatives.

Design and Application Research of Audit Risk Control Program

Strict procedures for prevention and control of audit risk is one of the most effective measures, so the research on design of audit risk control standardization program has very great practical application value. Combined with the research needs, the CPA practices respectively utilize means of summarizing method, for example, analysis method and enumeration method. The CPA practice also revealing the audit risk and audit report risk associated with different basis, combining with the traditional audit risk model. The specific content concludes the report of audit risk and then designs including audit reports preparation, audit draft, final review report to track a feedback of visit to the control standard operating requirements, and then puts forward the independent auditing standards. The auditing standards should be added to the proposal of audit report program.