scholarly journals Mitigating bias from intermittent measurement of time‐dependent covariates in failure time analysis

2020 ◽  
Vol 39 (13) ◽  
pp. 1833-1845 ◽  
Author(s):  
Shu Jiang ◽  
Richard J. Cook ◽  
Leilei Zeng
Keyword(s):  
Failure Time ◽  
Time Dependent ◽  
Time Analysis ◽  
Time Dependent Covariates ◽  
Intermittent Measurement

ACCELERATED FAILURE TIME MODELS COMPARISON TO THE PROPORTIONAL HAZARD MODEL FOR TIME-DEPENDENT COVARIATES WITH RECURRING EVENTS

2014 ◽  
Vol 21 (02) ◽  
pp. 1450010 ◽  
Author(s):  
ALEXANDRE C. MENDES ◽  
NASSER FARD
Keyword(s):  
Failure Time ◽  
Hazard Model ◽  
Parametric Model ◽  
Time Dependent ◽  
Accelerated Failure Time ◽  
Proportional Hazard ◽  
Proportional Hazard Model ◽  
Accelerated Failure Time Models ◽  
Accelerated Failure ◽  
Time Dependent Covariates

This paper presents an analysis of parametric survival models and compares their applications to time to event data used to validate the approximation for repeated events applying the Proportional Hazard Model (PHM) proposed in Mendes and Fard [Int. J. Reliab., Qual. Saf. Eng.19(6) (2012) 1240004.1–1240004.18]. The subjects studied do not show degrading failures, allowing the comparison between accelerated failure time models with the PHM. Results showed the applicability of the Weibull model and the versatility of the PHM not only to match the results of the parametric model, but also to allow the implementation of time-dependent covariates, resulting in superior model fit and more insightful interpretation for the covariate hazards. The paper contribution is to present the PHM as a simpler, more robust model to determine the acceleration factor for reliability testing when compared to the formidable task of fitting a parametric model for the distribution of failure. The Kaplan–Meier method may provide misleading guidance for covariate significance when time-dependent covariates are applied; however, relevant graphical screening is supplied. Notwithstanding, the PHM provides additional options to treat the repeated observations applying robust covariance correction for lack of heterogeneity in the fixed effects model or adopting the stratified model that absorbs the error using the stratification concept.

Genetic evaluation of productive life in Iranian Holstein using survival analysis

2005 ◽  
Vol 2005 ◽  
pp. 131-131 ◽  
Author(s):  
M. Dadpasand Taromsari ◽  
S. R. Miraei-Ashtiani ◽  
M. Moradi Shahrebabak ◽  
R. Vaez Torshizi
Keyword(s):  
Survival Analysis ◽  
Dairy Cattle ◽  
Genetic Parameters ◽  
Linear Models ◽  
Failure Time ◽  
Genetic Evaluation ◽  
Time Analysis ◽  
Productive Life ◽  
Analysis Methodology ◽  
Time Dependent Covariates

Improvement of herd life increases profitability due to lower replacement costs of heifers, higher proportion of mature cows that produce at their maximum potential and increased opportunity for voluntary culling. Functional productive life (PL) after adjustment for production is the ability of a cow to remain healthy and delay involuntary culling (Ducrocq et al 1988). Survival or failure time analysis has replaced linear model approaches for routine genetic evaluation of dairy cattle in several countries (Sewalem et. al. 2003). It allows proper treatment of censored data, inclusion of time-dependent covariates and skewed or non normal distribution of data. Approximate estimates of the heritability of longevity traits typically range from 0.05 to 0.10 and 0.15 to 0.20 using linear models and survival analysis, respectively (Vollema et al. 2000 and Caraviello et. al. 2004). The objective of this study was to apply survival analysis methodology for assessing the most important factor influencing PL, estimation of genetic parameters of productive life and genetic evaluation in Iranian Holstein dairy cattle.

scholarly journals Marginal quantile regression for longitudinal data analysis in the presence of time-dependent covariates

2020 ◽  
Vol 0 (0) ◽  
Author(s):  
I-Chen Chen ◽  
Philip M. Westgate
Keyword(s):  
Parameter Estimation ◽  
Longitudinal Data ◽  
Quantile Regression ◽  
Mean Squared Error ◽  
Time Dependent ◽  
Regression Methods ◽  
Squared Error ◽  
Highly Correlated ◽  
Time Dependent Covariates ◽  
Independence Structure

AbstractWhen observations are correlated, modeling the within-subject correlation structure using quantile regression for longitudinal data can be difficult unless a working independence structure is utilized. Although this approach ensures consistent estimators of the regression coefficients, it may result in less efficient regression parameter estimation when data are highly correlated. Therefore, several marginal quantile regression methods have been proposed to improve parameter estimation. In a longitudinal study some of the covariates may change their values over time, and the topic of time-dependent covariate has not been explored in the marginal quantile literature. As a result, we propose an approach for marginal quantile regression in the presence of time-dependent covariates, which includes a strategy to select a working type of time-dependency. In this manuscript, we demonstrate that our proposed method has the potential to improve power relative to the independence estimating equations approach due to the reduction of mean squared error.

Sign in / Sign up

Export Citation Format

Share Document