scholarly journals K-sample omnibus non-proportional hazards tests based on right-censored data

2020 ◽  
Vol 29 (10) ◽  
pp. 2830-2850
Author(s):  
Malka Gorfine ◽  
Matan Schlesinger ◽  
Li Hsu
Keyword(s):  
Proportional Hazards ◽  
R Package ◽  
Right Censoring ◽  
Sample Space ◽  
Right Censored Data ◽  
Hazard Functions ◽  
Space Partition ◽  
Higher Power ◽  
Event Times ◽  
Partition Tests

This work presents novel and powerful tests for comparing non-proportional hazard functions, based on sample–space partitions. Right censoring introduces two major difficulties, which make the existing sample–space partition tests for uncensored data non-applicable: (i) the actual event times of censored observations are unknown and (ii) the standard permutation procedure is invalid in case the censoring distributions of the groups are unequal. We overcome these two obstacles, introduce invariant tests, and prove their consistency. Extensive simulations reveal that under non-proportional alternatives, the proposed tests are often of higher power compared with existing popular tests for non-proportional hazards. Efficient implementation of our tests is available in the R package KONPsurv, which can be freely downloaded from CRAN.

scholarly journals Transformation model estimation of survival under dependent truncation and independent censoring

2018 ◽  
Vol 28 (12) ◽  
pp. 3785-3798 ◽  
Author(s):  
Sy Han Chiou ◽  
Matthew D Austin ◽  
Jing Qian ◽  
Rebecca A Betensky
Keyword(s):  
R Package ◽  
Right Censoring ◽  
Model Fit ◽  
Transformation Model ◽  
Source Population ◽  
Event Time ◽  
Incubation Study ◽  
Regression Diagnostic ◽  
Event Times ◽  
Model Approach

Truncation is a mechanism that permits observation of selected subjects from a source population; subjects are excluded if their event times are not contained within subject-specific intervals. Standard survival analysis methods for estimation of the distribution of the event time require quasi-independence of failure and truncation. When quasi-independence does not hold, alternative estimation procedures are required; currently, there is a copula model approach that makes strong modeling assumptions, and a transformation model approach that does not allow for right censoring. We extend the transformation model approach to accommodate right censoring. We propose a regression diagnostic for assessment of model fit. We evaluate the proposed transformation model in simulations and apply it to the National Alzheimer’s Coordinating Centers autopsy cohort study, and an AIDS incubation study. Our methods are publicly available in an R package, tranSurv.

The asymptotic distribution of the Net Benefit estimator in presence of right-censoring

2021 ◽  
pp. 096228022110370
Author(s):  
Brice Ozenne ◽  
Esben Budtz-Jørgensen ◽  
Julien Péron
Keyword(s):  
Asymptotic Distribution ◽  
Nuisance Parameter ◽  
Real Data ◽  
R Package ◽  
Right Censoring ◽  
Drop Out ◽  
Net Benefit ◽  
Asymptotic Results ◽  
Finite Samples ◽  
Benefit Risk Assessment

The benefit–risk balance is a critical information when evaluating a new treatment. The Net Benefit has been proposed as a metric for the benefit–risk assessment, and applied in oncology to simultaneously consider gains in survival and possible side effects of chemotherapies. With complete data, one can construct a U-statistic estimator for the Net Benefit and obtain its asymptotic distribution using standard results of the U-statistic theory. However, real data is often subject to right-censoring, e.g. patient drop-out in clinical trials. It is then possible to estimate the Net Benefit using a modified U-statistic, which involves the survival time. The latter can be seen as a nuisance parameter affecting the asymptotic distribution of the Net Benefit estimator. We present here how existing asymptotic results on U-statistics can be applied to estimate the distribution of the net benefit estimator, and assess their validity in finite samples. The methodology generalizes to other statistics obtained using generalized pairwise comparisons, such as the win ratio. It is implemented in the R package BuyseTest (version 2.3.0 and later) available on Comprehensive R Archive Network.

scholarly journals Dependence modeling for recurrent event times subject to right‐censoring with D‐vine copulas

Biometrics ◽  
2019 ◽  
Vol 75 (2) ◽  
pp. 439-451 ◽  
Author(s):  
Nicole Barthel ◽  
Candida Geerdens ◽  
Claudia Czado ◽  
Paul Janssen
Keyword(s):  
Right Censoring ◽  
Recurrent Event ◽  
Dependence Modeling ◽  
Event Times ◽  
Vine Copulas

Estimation from cross-sectional data under a semiparametric truncation model

Biometrika ◽  
2020 ◽  
Vol 107 (2) ◽  
pp. 449-465
Author(s):  
C Heuchenne ◽  
J De Uña-Álvarez ◽  
G Laurent
Keyword(s):  
Mean Squared Error ◽  
Disease Onset ◽  
Lifetime Distribution ◽  
Parametric Models ◽  
Parametric Family ◽  
Cross Sectional ◽  
Right Censored Data ◽  
Lifetime Distributions ◽  
Likelihood Approach ◽  
Event Times

Summary Cross-sectional sampling is often used when investigating inter-event times, resulting in left-truncated and right-censored data. In this paper, we consider a semiparametric truncation model in which the truncating variable is assumed to belong to a certain parametric family. We examine two methods of estimating both the truncation and the lifetime distributions. We obtain asymptotic representations of the estimators for the lifetime distribution and establish their weak convergence. Both of the proposed estimators perform better than Wang’s (1991) nonparametric maximum likelihood estimator in terms of the integrated mean squared error, when the parametric family for the truncation is sufficiently close to its true distribution. The full likelihood approach is preferable to the conditional likelihood approach in estimating the lifetime distribution, though not necessarily the truncation distribution. In an application to Alzheimer’s disease data, hypothesis tests reject the uniform truncation distribution, but several other parametric models lead to similar behaviour of the truncation and lifetime distributions after disease onset.

scholarly journals MiRKAT: kernel machine regression-based global association tests for the microbiome

2020 ◽  
Author(s):  
Nehemiah Wilson ◽  
Ni Zhao ◽  
Xiang Zhan ◽  
Hyunwook Koh ◽  
Weijia Fu ◽  
...  
Keyword(s):  
R Package ◽  
Effect Sizes ◽  
Supplementary Information ◽  
Time To Event ◽  
Kernel Machine ◽  
Association Testing ◽  
Higher Power ◽  
Kernel Machine Regression ◽  
Two Measures ◽  
Rv Coefficient

Abstract Summary Distance-based tests of microbiome beta diversity are an integral part of many microbiome analyses. MiRKAT enables distance-based association testing with a wide variety of outcome types, including continuous, binary, censored time-to-event, multivariate, correlated and high-dimensional outcomes. Omnibus tests allow simultaneous consideration of multiple distance and dissimilarity measures, providing higher power across a range of simulation scenarios. Two measures of effect size, a modified R-squared coefficient and a kernel RV coefficient, are incorporated to allow comparison of effect sizes across multiple kernels. Availability and implementation MiRKAT is available on CRAN as an R package. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals ESTIMATING PANEL DATA DURATION MODELS WITH CENSORED DATA

2008 ◽  
Vol 24 (5) ◽  
pp. 1254-1276 ◽  
Author(s):  
Sokbae Lee
Keyword(s):  
Panel Data ◽  
Fixed Effects ◽  
Proportional Hazards ◽  
Proportional Hazards Model ◽  
Unobserved Heterogeneity ◽  
Right Censoring ◽  
Duration Models ◽  
Finite Sample ◽  
Asymptotically Normal ◽  
Monte Carlo Studies

This paper presents a method for estimating a class of panel data duration models, under which an unknown transformation of the duration variable is linearly related to the observed explanatory variables and the unobserved heterogeneity (or frailty) with completely known error distributions. This class of duration models includes a panel data proportional hazards model with fixed effects. The proposed estimator is shown to be n1/2-consistent and asymptotically normal with dependent right censoring. The paper provides some discussions on extending the estimator to the cases of longer panels and multiple states. Some Monte Carlo studies are carried out to illustrate the finite-sample performance of the new estimator.

An improved model for the French genetic evaluation of dairy bulls on length of productive life of their daughters

2005 ◽  
Vol 80 (3) ◽  
pp. 249-256 ◽  
Author(s):  
V. Ducrocq
Keyword(s):  
Proportional Hazards ◽  
Proportional Hazards Model ◽  
Analysis Model ◽  
Genetic Trend ◽  
Hazard Functions ◽  
Stage Of Lactation ◽  
Dairy Bulls ◽  
Estimated Breeding Values ◽  
Improved Model ◽  
Efficiency Of Selection

AbstractFunctional longevity of dairy cows has been routinely evaluated in France since 1997 using a survival analysis model. Recently, we proposed a genetic trend validation test that could be used before including national data in an international evaluation of bulls on longevity of their daughters. Its application to the French Holstein data revealed a large overestimation of the genetic trend. It was found that the bias is the result of a change in the baseline hazard rate over time. A new proportional hazards model is proposed which accounts for this change. In the new model, the baseline is described as a stratified, piecewise Weibull hazard function within lactation, i.e. a function of the number of days since the most recent calving. Stratification is within year and parity. Different Weibull hazard functions are used over four periods: 0 to 270 days, 271 to 380 days, 381 days to day when dried, dry period until the next calving. The non-genetic effects included in the model were slightly different from the previous one. In particular the interaction effects between the within herd-year class of production and lactation number × stage of lactation on the one hand and year-season were accounted for. The estimated genetic variance was smaller than with the old model. The new genetic trend is almost flat. An illustration of the efficiency of selection on the estimated breeding values for longevity is presented.

scholarly journals SUN-018 No Difference in Breastfeeding Rates in Women with Polycystic Ovary Syndrome

2020 ◽  
Vol 4 (Supplement_1) ◽  
Author(s):  
Leeann Bui ◽  
Laura Cooney ◽  
Jen Birstler
Keyword(s):  
Polycystic Ovary Syndrome ◽  
Proportional Hazards ◽  
Polycystic Ovary ◽  
Cox Proportional Hazards ◽  
Right Censoring ◽  
Self Report ◽  
Ovary Syndrome ◽  
Cross Sectional ◽  
Median Response ◽  
Breastfeeding Rates

Abstract No Difference in Breastfeeding Rates in Women with Polycystic Ovary Syndrome Objective: Women with PCOS have increased rates of obesity and gestational weight gain compared to women without PCOS, factors which are associated with decreased breastfeeding (BF). Thus, our objective was to evaluate if women with PCOS were less likely to initiate BF. Design: Cross-sectional analysis of participants in the PRAMS (Pregnancy Risk Assessment Monitoring System) dataset, a national questionnaire from the Centers for Disease Control (CDC) sent to postpartum mothers 2–9 months after delivery. PCOS status and BF were by self-report. Logistic regression was used to assess odds of ever BF. Length of BF was assessed using Cox proportional hazards with right censoring for women who were still BF at the time of follow-up. PRAMS complex survey design was accounted for. Results: PCOS status was available for 14 states. Median response time was 3.7 months postpartum. Data from 16,036 participants were included which represents 855,302 women due to sample weights. 6.6% of women reported having PCOS and 83.8% reported ever BF. Compared to women with a normal BMI, women who were overweight or obese had decreased odds of BF (OR: 0.7, 95% CI: 0.6–0.9, P=0.01; OR: 0.6, 95% CI: 0.5–0.7, P<0.001 respectively); however, PCOS was not associated with BF (OR: 1.1, 95% CI: 0.9–1.3, P=0.6). In multivariate analysis, women with PCOS still were at no decreased odds of BF after adjusting for age, BMI, race, ethnicity, infertility treatment, and delivery factors (ORadj:1.1; 95% CI: 0.8–1.4; P=0.6). Variables associated with decreased odds of BF included: overweight/obesity, age ≤ 19 yrs (vs. 25–29), Black race, smoking, undesired pregnancy intent, gestational age ≤27 wks, and prior live birth. Variables associated with increased BF included: age 30–39 yrs, hospital stay 1–2 days (vs. 3–5), Hispanic ethnicity, and ≥ 3 life stressors. In multivariable Cox models, women with PCOS did not have a shorter length of BF (HRadj: 0.9, 95% CI: 0.8–1.1, P=0.3). Conclusion: Given the rise of the national rates of obesity and clear maternal and neonatal benefits to breastfeeding, understanding the predictors of BF success is paramount. In this national survey, women with PCOS were at no decreased odds of BF, despite confirming the association between overweight/obesity and decreased BF. However, our data still supports the clinical relevance of carefully targeting women with PCOS for BF education due to the association of PCOS with increased BMI. Additional prospective studies are needed to fully understand the association between PCOS and BF.

scholarly journals The Generalized Gamma Shared Frailty Model under Different Baseline Distributions

2019 ◽  
Vol 4 (1) ◽  
pp. 219-231
Author(s):  
Sukhmani Sidhu ◽  
Kanchan Jain ◽  
Suresh K. Sharma Sharma
Keyword(s):  
Survival Data ◽  
Bayes Factor ◽  
Real Life ◽  
Frailty Model ◽  
Information Criteria ◽  
Estimation Methods ◽  
Hazard Functions ◽  
Generalized Gamma ◽  
Shared Frailty ◽  
Event Times

In the analysis of clustered survival data, shared frailty models are often used when observations in the same group share common unknown risk factors or frailty. There is dependence in the event times belonging to the same group, while event times from different groups are conditionally independent given their covariates. In such models, the known effect on survival time is described using the baseline distribution and regression coefficients while the unknown effect is described through a frailty distribution. In this paper, the Gompertz, log-logistic, and generalized exponential distributions are studied as baseline distributions, under a shared frailty effect described by the generalized gamma distribution. Their hazard functions have been compared and their applicability under different settings and performance with generalized gamma frailty has been explored. These models are fitted to three real life datasets using Bayesian estimation methods and compared using the Bayesian Information Criteria (AIC, BIC, and DIC) and the Bayes Factor.

Sign in / Sign up

Export Citation Format

Share Document