Derive Risk Estimates Research Articles

Ambient Anthropogenic Carbons and Pediatric Respiratory Infections: A Case-Crossover Analysis in the Megacity Beijing.

Carbon loading in airway cells has shown to worsen function of antimicrobial peptides, permitting increased survival of pathogens in the respiratory tract; however, data on the impacts of carbon particles on childhood acute respiratory infection (ARI) is limited. We assembled daily health data on outpatient visits for ARI (bronchitis, pneumonia, and total upper respiratory infection [TURI]) in children aged 0-14years between 2015 and 2019 in Beijing, China. Anthropogenic carbons, including black carbon (BC) and its emission sources, and wood smoke particles (delta carbon, ultra-violet absorbing particulate matter, and brown carbon) were continuously monitored. Using a time-stratified case-crossover approach, conditional logistic regression was performed to derive risk estimates for each outcome. A total of 856,899 children were included, and a wide range of daily carbon particle concentrations was observed, with large variations for BC (0.36-20.44) and delta carbon (0.48-57.66μg/m3). Exposure to these particles were independently associated with ARI, with nearly linear exposure-response relationships. Interquartile range increases in concentrations of BC and delta carbon over prior 0-8days, we observed elevation of the odd ratio of bronchitis by 1.201 (95% confidence interval, 1.180, 1.221) and 1.048 (95% CI, 1.039, 1.057), respectively. Stronger association was observed for BC from traffic sources, which increased the odd ratio of bronchitis by 1.298 (95% CI, 1.273, 1.324). Carbon particles were also associated with elevated risks of pneumonia and TURI, and subgroup analyses indicated greater risks among children older than 6years. Our findings suggested that anthropogenic carbons in metropolitan areas may pose a significant threat to clinical manifestations of respiratory infections in vulnerable populations.

Floods and droughts: a multivariate perspective

Abstract. Multivariate or compound hydrological-extreme events such as successive floods, large-scale droughts, or consecutive drought-to-flood events challenge water management and can be particularly impactful. Still, the multivariate nature of floods and droughts is often ignored by studying individual characteristics only, which can lead to the under- or overestimation of risk. Studying multivariate extremes is challenging because of variable dependencies and because they are even less abundant in observational records than univariate extremes. In this review, I discuss different types of multivariate hydrological extremes and their dependencies, including regional extremes affecting multiple locations, such as spatially connected flood events; consecutive extremes occurring in close temporal succession, such as successive droughts; extremes characterized by multiple characteristics, such as floods with jointly high peak discharge and flood volume; and transitions between different types of extremes, such as drought-to-flood transitions. I present different strategies to describe and model multivariate extremes and to assess their hazard potential, including descriptors of multivariate extremes, multivariate distributions and return periods, and stochastic and large-ensemble simulation approaches. The strategies discussed enable a multivariate perspective on hydrological extremes, which allows us to derive risk estimates for extreme events described by more than one variable.

Development of a prognostic prediction model to estimate the risk of multiple chronic diseases

IntroductionThe ability to estimate risk of multimorbidity will provide valuable information to patients and primary care practitioners in their preventative efforts. Current methods for prognostic prediction modelling are insufficient for the estimation of risk for multiple outcomes, as they do not properly capture the dependence that exists between outcomes. ObjectivesWe developed a multivariate prognostic prediction model for the 5-year risk of diabetes, hypertension, and osteoarthritis that quantifies and accounts for the dependence between each disease using a copula-based model. MethodsWe used data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) from 2009 onwards, a collection of electronic medical records submitted by participating primary care practitioners across Canada. We identified patients 18 years and older without all three outcome diseases and observed any incident diabetes, osteoarthritis, or hypertension within 5-years, resulting in a large retrospective cohort for model development and internal validation (n=425228). First, we quantified the dependence between outcomes using unadjusted and adjusted ϕ coefficients. We then estimated a copula-based model to quantify the non-linear dependence between outcomes that can be used to derive risk estimates for each outcome, accounting for the observed dependence. Copula-based models are defined by univariate models for each outcome and a dependence function, specified by the parameter θ. Logistic regression was used for the univariate models and the Frank copula was selected as the dependence function. ResultsAll outcome pairs demonstrated statistically significant dependence that was reduced after adjusting for covariates. The copula-based model yielded statistically significant θ parameters in agreement with the adjusted and unadjusted ϕ coefficients. Our copula-based model can effectively be used to estimate trivariate probabilities. DiscussionQuantitative estimates of multimorbidity risk inform discussions between patients and their primary care practitioners around prevention in an effort to reduce the incidence of multimorbidity.

Development of a prognostic prediction model to estimate the risk of multiple chronic diseases: constructing a copula-based model using Canadian primary care electronic medical record data.

IntroductionThe ability to estimate risk of multimorbidity will provide valuable information to patients and primary care practitioners in their preventative efforts. Current methods for prognostic prediction modelling are insufficient for the estimation of risk for multiple outcomes, as they do not properly capture the dependence that exists between outcomes.ObjectivesWe developed a multivariate prognostic prediction model for the 5-year risk of diabetes, hypertension, and osteoarthritis that quantifies and accounts for the dependence between each disease using a copula-based model.MethodsWe used data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) from 2009 onwards, a collection of electronic medical records submitted by participating primary care practitioners across Canada. We identified patients 18 years and older without all three outcome diseases and observed any incident diabetes, osteoarthritis, or hypertension within 5-years, resulting in a large retrospective cohort for model development and internal validation (n=425,228). First, we quantified the dependence between outcomes using unadjusted and adjusted Ø coefficients. We then estimated a copula-based model to quantify the non-linear dependence between outcomes that can be used to derive risk estimates for each outcome, accounting for the observed dependence. Copula-based models are defined by univariate models for each outcome and a dependence function, specified by the parameter θ. Logistic regression was used for the univariate models and the Frank copula was selected as the dependence function.ResultsAll outcome pairs demonstrated statistically significant dependence that was reduced after adjusting for covariates. The copula-based model yielded statistically significant θ parameters in agreement with the adjusted and unadjusted Ø coefficients. Our copula-based model can effectively be used to estimate trivariate probabilities.DiscussionQuantitative estimates of multimorbidity risk inform discussions between patients and their primary care practitioners around prevention in an effort to reduce the incidence of multimorbidity.

The Basel II internal ratings based (IRB) model and the transition impact on the listed Greek banks

The Basel II accord implemented in 2006, meant that banks worldwide could use Internal Ratings-Based (IRB) models, in order to evaluate the components of their Capital Adequacy Ratio (CAR). In 2017 the IRB approach was also included into the Basel III Framework. The financial crisis of 2007-09 revealed the unsustainability of the Greek debt and lead the economy into a deep and prolonged recession. As a result the Greek stock exchange and the quotes of Greek banks plummeted. Five out of the six listed Greek banks adopted IRB models between 2008 and 2017. This paper investigates whether Greek IRB-banks performed better, within the adverse economic conditions, compared to the non-IRB banks. Difference-in-difference (DiD) and spatial DiD methodologies are employed and annual data from the Athens stock exchange, over the period 2001–2017, is used. It appears that there is a negative impact of the IRB implementation on listed Greek banks. This is mainly attributed to the higher cost involved in deriving risk estimates with the IRB approach, especially during this period of the stressful economic conditions. It is also attributed though to the restrictive regulatory measures imposed on Greek banks, which minimized any benefits derived from the IRB transition.

Predicting severe clinical events by learning about life-saving actions and outcomes using distant supervision.

Medical error is a leading cause of patient death in the United States. Among the different types of medical errors, harm to patients caused by doctors missing early signs of deterioration is especially challenging to address due to the heterogeneity of patients' physiological patterns. In this study, we implemented risk prediction models using the gradient boosted tree method to derive risk estimates for acute onset diseases in the near future. The prediction model uses physiological variables as input signals and the time of the administration of outcome-related interventions and discharge diagnoses as labels. We examine four categories of acute onset illness: acute heart failure (AHF), acute lung injury (ALI), acute kidney injury (AKI), and acute liver failure (ALF). To develop and test the model, we consider data from two sources: 23,578 admissions to the Intensive Care Unit (ICU) from the MIMIC-3 dataset (Beth-Israel Hospital) and 16,612 ICU admissions on hospitals affiliated with our institution (University of Washington Medical Center and Harborview Medical Center, the UW-CDR dataset). We systematically identify outcome-related interventions for each acute organ failure, then use them, along with discharge diagnoses, to label proxy events to train gradient boosted trees. The trained models achieve the highest F1 score with a value of 0.6018 when predicting the need for life-saving interventions for ALI within the next 24h in the MIMIC-3 dataset while showing a median F1 score of 0.3850 from all acute organ failures in both datasets. The approach also achieves the highest F1 score of 0.6301 when classifying a patient's ALI status at the time of discharge from the MIMIC-3 dataset, with a median F1 score of 0.4307 in both datasets. This study shows the potential for using the time of outcome-related intervention administrations and discharge diagnoses as labels to train supervised machine learning models that predict the risk of acute onset illnesses.

DNA topoisomerase IIα and Ki67 are prognostic factors in patients with hepatocellular carcinoma

The present study was designed to determine the significance of DNA topoisomerase IIa (TopoIIα) and Ki67 in hepatocellular carcinoma cells (HCCs). The present study included 353 patients with HCC. The association of clinicopathological data with the expression of TopoIIα and Ki67 by immunohistochemistry was analyzed by χ2 test. Cox multivariate proportional hazards regression analysis and Kaplan-Meier analysis were performed with all the variables to derive risk estimates associated with overall survival (OS)/recurrence-free survival (RFS) and to control for confounders. TopoIIα and Ki67 were detected in the nuclei of the tumor cells. With TopoIIα, 35.7% of cells exhibited high expression, which was associated with tumor-node-metastasis stage, tumor size and α-fetoprotein level. With Ki67, 37.1% of cells exhibited high expression, which was associated with tumor-node-metastasis stage, tumor size and α-fetoprotein level. Correlation was identified between the expression level of TopoIIα and Ki67 in HCCs (r=0.444). Multivariate analysis revealed that high TopoIIα expression is a prognostic indicator for RFS [hazard ratio (HR), 2.002; 95% confidence interval (CI), 1.429–2.806] and OS (HR, 2.749; 95% CI, 1.919–3.939), and high Ki67 expression is a prognostic indicator for OS (HR, 1.816; 95% CI, 1.273–2.589). The TopoIIα-low group had a significantly increased RFS rate (55.6 vs. 31.7%) and OS rate (66.5 vs. 23.8%) compared with the TopoIIα-high group. The OS rate was increased in the Ki67-low group compared with the Ki67-high group (67.0 vs. 26.5%). Expression of TopoIIα and Ki67 are independent prognostic factors for survival in HCCs. TopoIIα was positively associated with Ki67 expression.

Application of the ICRP respiratory tract model to estimate pulmonary retention of industrially sampled indium-containing dusts

Inhalation of indium-containing dusts is associated with the development of indium lung disease. Workers may be exposed to several different chemical forms of indium; however, their lung dosimetry is not fully understood. We characterized the physicochemical properties and measured the lung dissolution kinetics of eight indium-containing dusts. Indium dissolution rates in artificial lung fluids spanned two orders of magnitude. We used the International Commission on Radiological Protection (ICRP) human respiratory model (HRTM) to estimate pulmonary indium deposition, retention and biokinetic clearance to blood. For a two-year (median workforce tenure at facility) exposure to respirable-sized particles of the indium materials, modeled indium clearance (>99.99% removed) from the alveolar-interstitial compartment was slow for all dusts; salts would clear in 4 years, sintered indium–tin oxide (ITO) would clear in 9 years, and indium oxide would require 48 years. For this scenario, the ICRP HRTM predicted that indium translocated to blood would be present in that compartment for 3.5–18 years after cessation of exposure, depending on the chemical form. For a 40-year exposure (working lifetime), clearance from the alveolar–interstitial compartment would require 5, 10 and 60 years for indium salts, sintered ITO and indium oxide, respectively and indium would be present in blood for 5–53 years after exposure. Consideration of differences in chemical forms of indium, dissolution rates, alveolar clearance and residence time in blood should be included in exposure assessment and epidemiological studies that rely on measures of total indium in air or blood to derive risk estimates.

Neuron-glial antigen 2 overexpression in hepatocellular carcinoma predicts poor prognosis.

To investigate whether neuron-glial antigen 2 (NG2) could be an effective prognostic marker in hepatocellular carcinoma (HCC). NG2 expression was semi-quantitatively scored from the immunohistochemistry (IHC) data based on the number of positive cells and the staining intensity. A total of 132 HCC specimens and 96 adjacent noncancerous tissue samples were analyzed by IHC for NG2 protein expression. To confirm the NG2 expression levels observed by IHC, we measured NG2 expression in 30 randomly selected tumor and adjacent noncancerous tissue samples by quantitative real-time polymerase chain reaction and Western blot. The correlations between NG2 protein expression and the clinicopathological features of HCC patients were analyzed using the χ (2) test. To assess the prognostic value of NG2 for HCC, the association between NG2 expression and survival was analyzed using the Kaplan-Meier method with the log-rank test. To further evaluate the prognostic value of NG2 expression, a Cox multivariate proportional hazards regression analysis was performed with all the variables to derive risk estimates related to disease-free and overall survival and to control for confounders. High NG2 expression was observed in significantly more primary tumor samples (63.6%; 84/132) compared with the adjacent noncancerous tissue samples (28.1%; 27/96) (P < 0.0001). Moreover, high NG2 protein expression was closely associated with tumor differentiation (χ (2) = 9.436, P = 0.0089), recurrence (χ (2) = 5.769, P = 0.0163), tumor-node-metastasis (TNM) stage (χ (2) = 8.976, P = 0.0027), and invasion (χ (2) = 5.476, P = 0.0193). However, no significant relationship was observed between NG2 protein expression in HCC and other parameters, such as age, sex, tumor size, serum alpha fetoprotein (AFP), tumor number, or tumor capsule. The log-rank test indicated a significant difference in the overall survival of HCC patients with high NG2 expression compared with those with low NG2 expression (29.2% vs 9.5%, P < 0.001). Moreover, NG2 expression in HCC tissue significantly correlated with disease-free survival (15.2% vs 6.7%, P < 0.001). Multivariate analysis showed that NG2 expression (HR = 2.035, P = 0.002), serum AFP (HR = 1.903, P = 0.003), TNM stage (HR = 2.039, P = 0.001), and portal vein invasion (HR = 1.938, P = 0.002) were independent prognostic indicators for OS in HCC patients. Furthermore, NG2 expression (HR = 1.974, P = 0.003), serum AFP (HR = 1.767, P = 0.008), TNM stage (HR = 2.078, P = 0.001), tumor capsule (HR = 0.652, P = 0.045), and portal vein invasion (HR = 1.941, P = 0.002) were independent prognostic indicators for DFS in HCC patients. The up-regulation of NG2 is associated with poor prognosis in HCC. Therefore, NG2 could be useful as an additional prognostic marker to increase the resolution of traditional approaches.

Radiation-induced cancer: a modern view

Diagnostic medical radiation has been the most rapidly increasing component of population background radiation exposure in Western countries over the past decade. This trend is set to increase as CT scanning is readily available with burgeoning use in everyday clinical practice. Consequently, the issue of cancer induction from the doses received during diagnostic medical exposures is highly relevant. In this review we explain current understanding of potential cancer induction at low doses of sparsely ionising radiation. For cancers that may be induced at low doses, a mechanistic description of radiation-induced cancer is discussed, which, in combination with extrapolation of data based on population cohort studies, provides the basis of the currently accepted linear no-threshold model. We explore the assumptions made in deriving risk estimates, the controversies surrounding the linear no-threshold model and the potential future challenges facing clinicians and policy-makers with regards to diagnostic medical radiation and cancer risk, most notably the uncertainties regarding deriving risk estimates from epidemiological data at low doses.

Estimates of the potential risk of radiation-related cancer from screening in the UK

Estimates of the potential risk of radiation-related cancer from screening in the UK

Image mining utilising multitemporal image fusion based on map algebras and stereology to derive risk estimates and rates of change

This article extends the notion of image mining with map algebras and stereology. It considers natural or man-made spatial processes observable at the earth surface. Image mining addresses the chain from object identification from remote-sensing images through modelling, tracking on a series of images in time, prediction, towards communication to stakeholders. Image mining may benefit from image fusion during the object identification step and the tracking step. A map algebra when applied as a quantitative combination of different images, e.g. created at different times, may give cumulative information in space and time. Stereology allows us to assess basic properties from a series of images in a statistical way. Stereology is considered at the object level where the size of an increasing and decreasing object in space and time is addressed. In this article both are treated and applied from the decision-making perspective and they are illustrated with data on flooding. We conclude that map algebras and stereology are complementary in their role within an image-mining perspective.

Assessment of severity of the extreme River Tyne flood in January 2005 using gauged and historical information

The extreme Tyne (Northumbria, UK) flood in January 2005 provided the opportunity to reassess flood risk and to link peak discharge and flooded area to probability of occurrence. However, in spite of the UK guidance on flood risk assessment given in the Flood Estimation Handbook (FEH), there is still considerable subjectivity in deriving risk estimates. A particular problem for the Tyne arises from the effects of river bed gravel extraction both on the reliability of gauged discharges and in the interpretation of historical level data. In addition, attenuation and drawdown of Kielder Water has reduced downstream flood risk since 1982. Estimates from single-site, pooled estimates and historical information are compared. It is concluded that the return period of the observed flood was around 71 years on the lower Tyne and was probably the largest flood since 1815.

Are There Good Air Traffic Management Safety Indicators For Very Safe Systems?

European Air Traffic Management is extremely safe. The drawback to this safety record is that it is very difficult to estimate what the ‘underlying’ accident rate for mid-air collisions is now, or to detect any changes over time. The aim is to see if it possible to construct simple ATM safety indicators that correlate with this underlying accident rate. A perfect indicator would be simple to comprehend and capable of being calculated by a checklist process. This problem has been examined using a combination of analogies with simple ‘defensive’ systems with Markov process properties. An important concept is that of ‘system control’: the ability to determine the outcome against reasonably foreseen changes and variations of system parameters. The statistical distribution of future incidents has been analysed by focusing on an index – the Close Proximity Index (CPI) – of separation at the Closest Point of Approach. A promising indicator is ‘Incident Not Resolved by ATC’ (INRA) incidents in which the ground ATC defences have been ‘used up’. ATM Incidents can also be categorised in other ways. Two examples are reviewed: the risk-bearing category for Airproxes and ‘risk of collision/severity’ scores. The second is more promising conceptually, but the existing scoring system has not been demonstrated to have the properties necessary to derive risk estimates. The key question is: if someone says they know how to make a good estimate of the underlying accident rate, then how could this claim be tested? If it correlates very well with INRA, then what would be the argument for saying that it is a better indicator?

Venous thromboembolism associated with cyproterone acetate in combination with ethinyloestradiol (Dianette): observational studies using the UK General Practice Research Database.

To derive risk estimates for venous thromboembolism (VTE) in women prescribed cyproterone acetate combined with ethinyloestradiol (CPA/EE), a drug licensed in the UK for the treatment of women with acne or hirsutism. CPA/EE provides a treatment option for women with polycystic ovary syndrome (PCOS). CPA/EE has been associated with an increased risk of VTE. Using the General Practice Research Database, we conducted cohort and case-control analyses in all women aged 15-39 and then nested in a population of women of the same age with acne, hirsutism or PCOS. The incidence rate ratio (IRR) for VTE in women exposed to CPA/EE versus conventional combined oral contraceptives (COCs) was significantly raised (all women: 1.92; 95% CI: 1.22,2.88; nested: 2.51; 95% CI: 1.07,5.75). Using exposure to conventional COCs as the reference, the adjusted odds ratio (ORadj) for VTE associated with CPA/EE was 1.45 (95% CI: 0.80,2.64) in all women and 1.71 (95% CI: 0.31,9.49) in women with acne, hirsutism or PCOS. The risk of VTE associated with CPA/EE use does not differ significantly from that associated with the use of conventional COCs. These data are reassuring and together with knowledge of the risks associated with other treatments for acne, in particular, should influence prescribing practice.

Enhanced mutation and neoplastic transformation in human cells by 29 kVp relative to 200 kVp X rays indicating a strong dependence of RBE on photon energy.

The fundamental assumption implicit in the use of the atomic bomb survivor data to derive risk estimates for occupational and medical exposures is that the gamma rays of Hiroshima and Nagasaki are considered as equal efficiencies to other low LET radiations up to an LET of 10 keV.micron-1. For breast cancer induction, neoplastic cell transformation, mutation, reciprocal translocations and dicentrics in human lymphocytes, a strong and very similar dependence of the RBE values on photon energy or on LET is observed. Experimental data on mutation induction and neoplastic cell transformation in human cells show that 29 kVp X rays are by a factor of 4 and 3.4, respectively, more effective compared with 200 kVp X rays. These data are in excellent agreement with the data in the literature.

Residential radon exposure and lung cancer: variation in risk estimates using alternative exposure scenarios.

The most direct way to derive risk estimates for residential radon progeny exposure is through epidemiologic studies that examine the association between residential radon exposure and lung cancer. However, the National Research Council concluded that the inconsistency among prior residential radon case-control studies was largely a consequence of errors in radon dosimetry. This paper examines the impact of applying various epidemiologic dosimetry models for radon exposure assessment using a common data set from the Iowa Radon Lung Cancer Study (IRLCS). The IRLCS uniquely combined enhanced dosimetric techniques, individual mobility assessment, and expert histologic review to examine the relationship between cumulative radon exposure, smoking, and lung cancer. The a priori defined IRLCS radon-exposure model produced higher odds ratios than those methodologies that did not link the subject's retrospective mobility with multiple, spatially diverse radon concentrations. In addition, the smallest measurement errors were noted for the IRLCS exposure model. Risk estimates based solely on basement radon measurements generally exhibited the lowest risk estimates and the greatest measurement error. The findings indicate that the power of an epidemiologic study to detect an excess risk from residential radon exposure is enhanced by linking spatially disparate radon concentrations with the subject's retrospective mobility.

Commentary

Commentary

Analysis of thyroid cancer data from the Ukraine after 'Chernobyl' using a two-mutation carcinogenesis model.

The thyroid cancer data of children in the northern regions of the Ukraine after the reactor accident at Chernobyl were combined with thyroid dose measurements in the same regions and analysed using a two-mutation carcinogenesis model. The best fit was obtained for radiation acting as an initiating agent, i.e. on the first mutation of the model. The observed relatively high increase of thyroid cancer incidence after 1990 in children exposed to radiation released after the reactor accident could be ascribed to the high thyroid doses and the relatively low background thyroid cancer incidence in children. The maximum annual incidence is predicted to occur fairly soon after the reactor accident, i.e. about 10 years. For adults, the predicted relative increase of annual thyroid cancers is much lower than for children younger than 20 years. The modelling results are used to derive risk estimates for radiation-induced thyroid cancer. These risk estimates are dependent on age at exposure, follow-up time and the background thyroid cancer incidence. The calculated excess absolute risk for a population of all ages is about one-third of that currently used by ICRP, but for children the calculated absolute risks are about a factor of 3 higher than derived in other epidemiological studies. The model results indicate that the excess absolute radiation risk per unit dose for children is about the same as or a little lower than that for adults.

Retrospective Temporal and Spatial Mobility of Adult Iowa Women

Human exposure assessments require a linkage between toxicant concentrations in occupied spaces and the receptor's mobility pattern. Databases reporting distinct populations' mobility in various parts of the home, time outside the home, and time in another building are scarce. Temporal longitudinal trends in these mobility patterns for specific age and gender groups are nonexistent. This paper describes subgroup trends in the spatial and temporal mobility patterns within the home, outside the home, and in another building for 619 Iowa females that occupied the same home for at least 20 years. The study found that the mean time spent at home for the participants ranged from a low of 69.4% for the 50‐59 year age group to a high of 81.6% for the over 80‐year‐old age group. Participants who lived in either one‐ or two‐ story homes with basements spent the majority of their residential occupancy on the first story. Trends across age varied for other subgroups by number of children, education, and urbadrural status. Since all of these trends were nonlinear, they indicate that error exists when assuming a constant, such as a 75% home occupancy factor, which has been advocated by some researchers and agencies. In addition, while aggregate data, such as presented in this report, are more helpll in deriving risk estimates for population subgroups, they cannot supplant good individual‐level data for determining risks.