Analysis Of Censored Data Research Articles

A new regression model for the analysis of bimodal censored data: A comparison with random survival forest

A new regression model for the analysis of bimodal censored data: A comparison with random survival forest

Quantitative fecal source characterization of urban municipal storm sewer system outfall ‘wet’ and ‘dry’ weather discharges

Urban areas are built environments containing substantial amounts of impervious surfaces (e.g., streets, sidewalks, roof tops). These areas often include elaborately engineered drainage networks designed to collect, transport, and discharge untreated stormwater into local surface waters. When left uncontrolled, these discharges may contain unsafe levels of fecal waste from sources such as sanitary sewage and wildlife even under dry weather conditions. This study evaluates paired measurements of host-associated genetic markers (log10 copies per reaction) indicative of human (HF183/BacR287 and HumM2), ruminant (Rum2Bac), canine (DG3), and avian (GFD) fecal sources, 12-hour cumulative precipitation (mm), four catchment land use metrics determined by global information system (GIS) mapping, and Escherichia coli (MPN/100 ml) from seven municipal separate storm sewer system outfall locations situated at the southern portion of the Anacostia River Watershed (District of Columbia, U.S.A.). A total of 231 discharge samples were collected twice per month (n = 24 sampling days) and after rain events (n = 9) over a 13-month period. Approximately 50 % of samples (n = 116) were impaired, exceeding the local E. coli single sample maximum of 2.613 log10 MPN/100 ml. Genetic quality controls indicated the absence of amplification inhibition in 97.8 % of samples, however 14.7 % (n = 34) samples showed bias in DNA recovery. Of eligible samples, quantifiable levels were observed for avian (84.1 %), human (57.4 % for HF183/BacR287 and 40 % for HumM2), canine (46.7 %), and ruminant (15.9 %) host-associated genetic markers. Potential links between paired measurements are explored with a recently developed Bayesian qPCR censored data analysis approach. Findings indicate that human, pet, and urban wildlife all contribute to storm outfall discharge water quality in the District of Columbia, but pollutant source contributions vary based on ‘wet’ and ‘dry’ conditions and catchment land use, demonstrating that genetic-based fecal source identification methods combined with GIS land use mapping can complement routine E. coli monitoring to improve stormwater management in urban areas.

Genetic fecal source identification in urban streams impacted by municipal separate storm sewer system discharges.

Municipal stormwater systems are designed to collect, transport, and discharge precipitation from a defined catchment area into local surface waters. However, these discharges may contain unsafe levels of fecal waste. Paired measurements of Escherichia coli, precipitation, three land use metrics determined by geographic information system (GIS) mapping, and host-associated genetic markers indicative of human (HF183/BacR287 and HumM2), ruminant (Rum2Bac), dog (DG3), and avian (GFD) fecal sources were assessed in 231 urban stream samples impacted by two or more municipal stormwater outfalls. Receiving water samples were collected twice per month (n = 24) and after rain events (n = 9) from seven headwaters of the Anacostia River in the District of Columbia (United States) exhibiting a gradient of impervious surface, residential, and park surface areas. Almost 50% of stream samples (n = 103) were impaired, exceeding the local E. coli single sample maximum assessment level (410 MPN/100 ml). Fecal scores (average log10 copies per 100 ml) were determined to prioritize sites by pollution source and to evaluate potential links with land use, rainfall, and E. coli levels using a recently developed censored data analysis approach. Dog, ruminant, and avian fecal scores were almost always significantly increased after rain or when E. coli levels exceeded the local benchmark. Human fecal pollution trends showed the greatest variability with detections ranging from 9.1% to 96.7% across sites. Avian fecal scores exhibited the closest connection to land use, significantly increasing in catchments with larger residential areas after rain events (p = 0.038; R2 = 0.62). Overall, results demonstrate that combining genetic fecal source identification methods with GIS mapping complements routine E. coli monitoring to improve management of urban streams impacted by stormwater outfalls.

Ranking based variable selection for censored data using AFT models

Numerous variable selection techniques have been developed for complete high-dimensional data but very few of them for censored data. The techniques for complete data must be modified if censoring is present. In this paper, we consider the variable selection technique for accelerated failure time (AFT) models by extending the ranking-based variable selection (RBVS) algorithm and its iterative procedure as proposed in the work of Baranowski et al. through the Stute’s weighted least square technique. Simulation studies are conducted to demonstrate the performance of the proposed methods. We further illustrate the performance of this method with a mantle cell lymphoma microarray example. When there is no correlation among the covariates, the proposed method outperforms the iterative sure independence screening and stability selection methods in terms of overall performance for high-dimensional data. Real data analysis also suggests that the proposed method can be chosen for high-dimensional censored data analysis in parallel to other methods in the literature.

Simulation of inference test performance for minimum inhibitory concentration censored data

The estimation of the minimum inhibitory concentration is usually performed by a method of serial dilutions by a factor of 2, introducing the overestimation of antimicrobial efficacy, quantified by a simulation model that shows that the variability of the bias is higher for the standard deviation, being dependent on the metric distance to the values of the concentrations used. We use a methodological approach through modeling and simulation for the measurement error of physical variables with censored information, proposing a new inference method based on the calculation of the exact probability for the set of possible samples from nmeasurements that allows quantifying the p-value in one or two independent sample tests for the comparison of censored data means. Tests based on exact probability methods offer a reasonable solution for small sample sizes, with statistical power varying according to the hypothesis evaluated, providing insight into the limitations of censored data analysis and providing a tool for decision making in the diagnosis of antimicrobial efficacy.

Generalized Lindley Shared Frailty Based on Reversed Hazard Rate

Due to the unavailability of complete data in various circumstances in biological, epidemiological, and medical studies, the analysis of censored data is very common among practitioners. But the analysis of bivariate censored data is not a regular mechanism because it is not necessary to always have independent data. Observed and unobserved covariates affect the variables under study. So, heterogeneity is present in the data. Ignoring observed and unobserved covariates may have objectionable consequences. But it is not easy to find that whether there is any effect of the unobserved covariate or not. Shared frailty models are the viable choice to counter such scenarios. However, due to certain restrictions such as the identifiability condition and the requirement that their Laplace transform exists, finding a frailty distribution can be difficult. As a result, in this paper, we introduce a new frailty distribution generalized Lindley (GL) for reversed hazard rate (RHR) setup that outperforms the gamma frailty distribution. So, our main motive is to establish a new frailty distribution under the RHR setup. By assuming exponential Gumbel (EG) and generalized inverted exponential (GIE) baseline distributions, we propose a new class of shared frailty models based on RHR. We estimate the parameters in these frailty models and use the Bayesian paradigm of the Markov Chain Monte Carlo (MCMC) technique. Model selection criteria have been performed for the comparison of models. We analyze Australian twin data and suggest a better model.

Nonparametric Multiple Imputation of Left Censored Event Times in Analysis of Follow-up Data

In this paper, we consider analysis of follow-up data where each event time is either right censored, observed, left censored or left truncated. In the case of left censoring, the covariates measured at baseline are considered as missing. The work is motivated by data from the MORGAM Project, which explores the association between cardiovascular diseases and their classic and genetic risk factors. We propose a nonparametric multiple imputation (NPMI) approach where the left censored event times and the missing covariates are imputed in hot deck manner. The left truncation due to deaths prior to baseline is compensated by Lexis diagram imputation introduced in the paper. After imputation, the standard estimation methods for right censored survival data can be directly applied. The performance of the proposed imputation approach is studied with simulated and real world data. The results suggest that the NPMI is a flexible and reliable approach to the analysis of left and right censored data.

Failure modes based censored data analysis for repairable systems and its industrial perspective

Reliability analysis of repairable systems (RS) has always been a challenging task for the industries especially when the collected data is censored. In real-world situation, the industries, generally preserve all the system’s related information such as the number of failures, time between failures and their respective failure modes (FMs) for future reliability analysis. The problem arises when FM wise analysis is required to be done for RS along with other censoring criteria such as test termination, removing the operating system from the study etc. The literature provides various models to deal with different types of censored data but lacks in providing FM based censored data analysis technique for RS, when both preventive and corrective maintenances (PM and CM) are treated as imperfect. To bridge this gap, the paper develops a technique and virtual age models for RS which can simultaneously address FM wise analysis with right and multiply censoring data considering both the PM and CM as imperfect. The paper also develops likelihood function using proposed models for parameter estimation. The proposed technique and models are demonstrated with the help of a case study selected from aviation industry. The paper then highlights the applicability of proposed models to industries dealing with complex and critical RS in conducting failure modes and effects analysis (FMEA) and remaining useful life (RUL) estimation. The paper also brings out as to how the proposed models can be converted to the existing models with some modifications.

Evaluation of Parametric Method Performance for Left-Censored Data and Recommendation of Using for Covid-19 Data Analysis

bjectives: Left-censored data, which is commonly seen in clinical studies, are frequently encountered in the literature, especially in the fields of food, environment, microbiology, and biochemistry. In this study, the most appropriate distribution between the negatively skewed distributions for left-censored data in Parametric Inverse Hazard Models was tried to be determined. Methods: Within the scope of the study, firstly, the data were produced uncensored according to different parameters of each distribution. Then, simulation studies were carried out in different censorship rates (15%, 25% and 35%) and various sample sizes (1000, 2000 and 3000) in order to determine the most appropriate distribution. AIC, AICC, HQIC, and CAIC information criteria were employed to compare the distribution performances. Since it was not possible to study simulations of all possible scenarios, scenarios similar to each other were generally preferred over others. Results: In the simulation results, the most appropriate distributions to be used for left-censored data in Parametric Inverse Hazard Models were found as Generalized Inverse Weibull as well as Log-Logistic, Log-Normal, Inverse Normal and Gamma distributions. It was also detected that the Marshal-Olkin distribution revealed a superior performance compared to the Modified Weibull, Generalized Gamma, Gamma, and Flexible Weibull distributions. Log logistics distribution gave the most appropriate result among the analyzed distributions in the examination made with real data application. Conclusion: The use of censored data analysis in evaluations in terms of Covid-19 is quite additive, considering that more statistical evaluation will be needed in the next period of the epidemic. Improved estimates can be obtained with this approach, especially in Covid-19 data analysis. Keywords: Covid-19, left-censored, limit of detection, Parametric Inverse Hazard Models

How has it changed? A comparative field evaluation of bioretention infiltration and treatment performance post-construction and at maturity

A conventionally-designed bioretention cell, consisting of very sandy media meant to achieve high infiltration rates, was studied for its hydrologic and treatment performance over two monitoring periods: 2013–2014 (immediately post-construction) and 2017–2018 (4–5 years post-construction). Given the limited literature on mature performance of bioretention, the purpose of this study was to determine how the effluent from the bioretention cell and the hydrology changed over 5 years. The hydrologic performance was maintained 5 years post-construction, with median volume reductions of 100% in both monitoring periods, despite an underutilization of the soil volume. Using censored data analysis and a 95% confidence level, the results revealed that the effluent water quality in 2017–2018 was improved compared to 2013–2014 for some parameters, e.g., dissolved solids and nitrogen species, and was maintained for phosphorus, metals, and suspended solids. These results suggest that for a reliable assessment of bioretention cell treatment performance, it is recommended to wait for soil and plant establishment, e.g., 2 years after construction. Between the inlet and outlet of the bioretention cell (2017–2018 data only), concentration decreased for nitrogen species and suspended solids, but did not significantly increase or decrease for alkalinity, hardness, dissolved solids, phosphorus and metals. Mass removal of all contaminants was very high, largely due to high volume reductions. Despite sustained hydrologic performance up to 5 years post-construction, there is a need for targeted bioretention design for enhanced treatment performance of dissolved contaminants.

Assessment of groundwater processes using censored data analysis incorporating non-detect chemical, physical, and biological data

In Europe most environmental based water quality research has focused on both nutrient and microbial contamination which can arise from agricultural processes and inadequate wastewater treatment. Recent work in Ireland has linked the presence of arsenic in groundwater at elevated concentrations at national and subnational scales with bedrock lithology serving as a strong predictor variable. Groundwater data was collected as part of an environmental impact assessment for a road construction project and this resulting groundwater geochemistry dataset was used in this present study to assess the geochemical controls of arsenic in natural waters in addition to biological and nutrient contamination. Physiochemical parameters, trace elements, nutrients, organics, and microbiological parameters were collected for every quarter for four years (2004–2008) in 67 wells. Due to differing sampling procedures and limitations in the data, only one quarter (November 2005) was used to understand groundwater geochemistry in greater detail. Multivariate statistical techniques were used to overcome the presence of non-detect data. This is an important consideration as while methods exist for chemical data, methods incorporating biological data are limited. Elevated levels of nitrate in groundwater may arise from the runoff of septic tanks and/or agricultural practices in the area. Both pesticides and polycyclic aromatic hydrocarbons were not detected in any wells signifying no anthropogenic contamination inputs. However, fuel products such as methyl tert-butyl ether were detected and potentially illustrate point source contamination, these were detected in only one well. Geochemical data indicate that elevated arsenic concentrations are present within alkali-oxic groundwaters through the desorption from Fe and Mn oxyhydroxides, i.e. alkali desorption. This study examines of the geochemistry of arsenic in groundwater in Ireland at a local scale. In addition, the multivariate methods used in this study were able to fully integrate both chemical and biological censored data, which may be applied in other regions with similar data.

The Lindley negative-binomial distribution: Properties, estimation and applications to lifetime data

Abstract A generalization of the Lindley distribution namely, Lindley negative-binomial distribution, is introduced. The Lindley and the exponentiated Lindley distributions are considered as sub-models of the proposed distribution. The proposed model has flexible density and hazard rate functions. The density function can be decreasing, right-skewed, left-skewed and approximately symmetric. The hazard rate function possesses various shapes including increasing, decreasing and bathtub. Furthermore, the survival and hazard rate functions have closed form representations which make this model tractable for censored data analysis. Some general properties of the proposed model are studied such as ordinary and incomplete moments, moment generating function, mean deviations, Lorenz and Bonferroni curve. The maximum likelihood and the Bayesian estimation methods are utilized to estimate the model parameters. In addition, a small simulation study is conducted in order to evaluate the performance of the estimation methods. Two real data sets are used to illustrate the applicability of the proposed model.

Sensitivity of censored data analysis to determine the characteristic value of structural timber

In structural timber tests, unintended failure mechanisms occur frequently in specimens and their results are called censored data. There are two censored data analysis: censored maximum likelihood estimation (CMLE) and Kaplan–Meier (KM) method. In this study, the precision of the censored data analysis was investigated to determine the characteristic value, 5th percentile value, of the structural timber. The results show that (1) the 5th percentile value was underestimated by ordinary data analysis methods; maximum likelihood estimation (MLE) and Order statistics. (2) CMLE with 30% lower tail censored data and KM method provided much more precise 5th percentile value. (3) The amount of under-measurement (5 MPa, 10 MPa, and 15 MPa in this simulation study) did not show significant effect on the 5th percentile determination in CMLE and KM method, but the proportion of censored data (percentage of unintended failure specimen; 10%, 20%, 30%, and 40%) affected the determination of 5th percentile value. (4) CMLE with 30% lower tail censored data and KM method showed good agreement in case that the data included unintended failure data up to 20%.

On Bayesian parameter estimation of beta log Weibull distribution under type-II censoring

This article presents analysis of Type-II censored data when the lifetime distribution follows beta log Weibull distribution. Parameter estimation under classical and Bayesian set up is undertaken. Asymptotic confidence intervals are obtained using observed information matrix. For comparison purpose bootstrap-p and bootstrap-t confidence intervals are also constructed. Bayes estimates of the unknown parameters and the reliability function are obtained using Tierney–Kadane’s approximation and Markov Chain Monte Carlo methods. A real data set is analyzed to illustrate the proposed strategies.

New approach for analysis of progressive Type-II censored data from the Pareto distribution

New approach for analysis of progressive Type-II censored data from the Pareto distribution

Bulk Lorentz factors of gamma-ray bursts

Knowledge of the bulk Lorentz factor Γ0 of gamma-ray bursts (GRBs) allows us to compute their comoving frame properties shedding light on their physics. Upon collisions with the circumburst matter, the fireball of a GRB starts to decelerate, producing a peak or a break (depending on the circumburst density profile) in the light curve of the afterglow. Considering all bursts with known redshift and with an early coverage of their emission, we find 67 GRBs (including one short event) with a peak in their optical or GeV light curves at a time tp. For another 106 GRBs we set an upper limit tpUL. The measure of tp provides the bulk Lorentz factor Γ0 of the fireball before deceleration. We show that tp is due to the dynamics of the fireball deceleration and not to the passage of a characteristic frequency of the synchrotron spectrum across the optical band. Considering the tp of 66 long GRBs and the 85 most constraining upper limits, we estimate Γ0 or a lower limit Γ0LL. Using censored data analysis methods, we reconstruct the most likely distribution of tp. All tp are larger than the time Tp,γ when the prompt γ-ray emission peaks, and are much larger than the time Tph when the fireball becomes transparent, that is, tp>Tp,γ>Tph. The reconstructed distribution of Γ0 has median value ~300 (150) for a uniform (wind) circumburst density profile. In the comoving frame, long GRBs have typical isotropic energy, luminosity, and peak energy ⟨ Eiso ⟩ = 3(8) × 1050 erg, ⟨ Liso ⟩ = 3(15) × 1047 erg s-1, and ⟨ Epeak ⟩ = 1(2) keV in the homogeneous (wind) case. We confirm that the significant correlations between Γ0 and the rest frame isotropic energy (Eiso), luminosity (Liso), and peak energy (Ep) are not due to selection effects. When combined, they lead to the observed Ep−Eiso and Ep−Liso correlations. Finally, assuming a typical opening angle of 5 degrees, we derive the distribution of the jet baryon loading which is centered around a few 10-6M⊙.

Semiparametric Bayesian analysis of high-dimensional censored outcome data

ABSTRACTThe Surveillance, Epidemiology and End Results (SEER) cancer database contains survival data for US individuals diagnosed with cancer. Semiparametric Bayesian methods are computationally expensive to fit for such large data-sets. This paper develops a cost-effective Markov chain Monte Carlo strategy for censored outcomes to fit a semiparametric bayesian analysis of SEER data of New Mexico. We use an accelerated failure time model, with Dirichlet process random effects for inter-subject variation, and intrinsic conditionally autoregressive random effects for spatial correlations. The results offer insights into differences in breast cancer mortality rates between ethnic groups, tumor grade and spatial effect of counties.

Bayesian Analysis of Censored Spatial Data Based on a Non-Gaussian Model

In this paper, we suggest using a skew Gaussian-log Gaussian model for the analysis of spatial censored data from a Bayesian point of view. This approach furnishes an extension of the skew log Gaussian model to accommodate to both skewness and heavy tails and also censored data. All of the characteristics mentioned are three pervasive features of spatial data. We utilize data augmentation method and Markov chain Monte Carlo (MCMC) algorithms to do posterior calculations. The methodology is illustrated using simulated data, as well as applying it to a real data set. Keywords: Censored data, data augmentation, non-Gaussian spatial models, outlier, unified skew Gaussian.

Semiparametric accelerated failure time model for the analysis of right censored data

The accelerated failure time model or accelerated life model relates the logarithm of the failure time linearly to the covariates. The parameters in the model provides a direct interpretation. In this paper, we review some newly developed practically useful estimation and inference methods for the model in the analysis of right censored data.

Ventricular Arrhythmia Burden in Patients With Heart Failure and Cardiac Resynchronization Devices: The Importance of Renal Function.

Chronic kidney disease (CKD) is a risk factor for arrhythmias in patients with heart failure (HF). However, the effects of CKD on ventricular arrhythmia (VA) burden in patients with cardiac resynchronization therapy and defibrillator (CRT-D) devices in a primary prevention setting are unknown. To determine whether baseline CKD is associated with increased risk of VA in patients implanted with primary prevention CRT-D devices. In this retrospective study, 199 consecutive primary prevention CRT-D recipients (2005-2010) were stratified by estimated glomerular filtration rate (eGFR) levels prior to device implantation with 106 (53.2%) ≥CKD III (eGFR < 60 mL/min/1.73 m2 ) (CKD group). CKD group patients were significantly older (70.0 ± 10 years vs. 61.3 ± 12 years, P < 0.05) with higher prevalence of ischemic cardiomyopathy (56.2% vs. 40.2%, P < 0.05). Detected ventricular tachycardia (VT)/ventricular fibrillation (VF) episodes resulting in device therapy occurred significantly more frequently in the CKD group [40/106(37.8%)] than controls [24/93(25.8%)], (odd ratio [OR] = 1.74, 95% confidence interval [CI] = 1.01-3.2, P = 0.05). At 5-year follow-up, interval censored data analysis showed 41% VT/VF incidence in the CKD group compared to 24% incidence in controls (P < 0.05). Cox proportional hazards model identified CKD > III as the only predictor of sustained VA in this group (adjusted hazard ratio [HR] 2.92, CI = 1.39-6.1, P = 0.004). Baseline CKD is a strong independent risk factor for VA in primary prevention CRT-D recipients. Further understanding of the underlying arrhythmogenic mechanisms relating to CKD may be of interest to allow appropriate correction and prevention. Device programming in this cohort may need to reflect this increased risk.