Empirical study of correlated survival times for recurrent events with proportional hazards margins and the effect of correlation and censoring

Villegas, Rodrigo; Julià, Olga; Ocaña, Jordi

doi:10.1186/1471-2288-13-95

Cited by 14 publications

(16 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Total time models, such as AG, PWP‐CP, WLW, and STC tend to produce larger estimated covariate effects. This may occur because total times within a patient may be highly correlated, resulting in a “carry‐over effect.” Such effects have been previously documented with the WLW model . We also observe that the coefficient estimates under the WLW model are larger than those from other competing models except the coefficient for family history.…”

Section: Analysis Of Cf Registry Datasupporting

confidence: 79%

On the self‐triggering Cox model for recurrent event data

Kim

Lin

Fine

2019

Statistics in Medicine

View full text Add to dashboard Cite

Recurrent event data frequently occur in longitudinal studies when subjects experience more than one event during the observation period. Often, the occurrence of subsequent events is associated with the experience of previous events.Such dependence is commonly ignored in the application of standard recurrent event methodology. In this paper, we utilize a Cox-type regression model with time-varying triggering effect depending on the number and timing of previous events to enhance both model fit and prediction. Parameter estimation and statistical inference is achieved via the partial likelihood. A statistical test procedure is provided to assess the existence of the triggering effects. We demonstrate our approach via comprehensive simulation studies and a real data analysis on chronic pseudomonas infections in young cystic fibrosis patients. Our model provides significantly better predictions than standard recurrent event models. KEYWORDSCox proportional hazards model, cystic fibrosis, recurrent event data 4240

show abstract

Section: Analysis Of Cf Registry Datasupporting

confidence: 79%

On the self‐triggering Cox model for recurrent event data

Kim

Lin

Fine

2019

Statistics in Medicine

View full text Add to dashboard Cite

show abstract

“…Discussions of the different approaches for recurrent models can be found in other articles [ 13 , 14 ]. Empirical comparison between different recurrent models suggested that results derived from different models could be substantially different, and the optimal model could depend on the context [ 15 ]. Models for multiple end-points in survival analysis include the bivariate survival model [ 16 ], the marginal model [ 8 ], and random effects/frailty models [ 17 ].…”

Section: Introductionmentioning

confidence: 99%

Comparison of risks of cardiovascular events in the elderly using standard survival analysis and multiple-events and recurrent-events methods

Efendi

Molenberghs

et al. 2015

BMC Med Res Methodol

View full text Add to dashboard Cite

BackgroundEpidemiological studies about cardiovascular diseases often rely on methods based on time-to-first-event for data analysis. Without taking into account multiple event-types and the recurrency of a specific cardiovascular event, this approach may underestimate the overall cardiovascular burden of some risk factors, if that is the goal of the study.MethodsIn this study we compare four different statistical approaches, all based on the Weibull distribution family of survival model, in analyzing cardiovascular risk factors. We use data from the Cardiovascular Health Study as illustration. The four models respectively are time-to-first-event only, recurrent-events only, multiple-event-types only, and joint recurrent and multiple-event-type models.ResultsAlthough the four models produce consistent results regarding the significance of the risk factors, the magnitude of the hazard ratios and their confidence intervals are different. The joint model produces hazard ratios that are substantially higher than the time-to-first-event model especially for the risk factors of smoking and diabetes.ConclusionOur findings suggest that for people with diabetes and are currently smoking, the overall cardiovascular burden of these risk factors would be substantially higher than that estimated using time-to-first-event method.Electronic supplementary materialThe online version of this article (doi:10.1186/s12874-015-0004-3) contains supplementary material, which is available to authorized users.

show abstract

“…Aside from the definition of the at‐risk variable, the partial likelihood is in this case identical to that of the AG model. Contrary to the AG model, however,

\overset{β}{true}

(as well as the event‐specific estimate

{\overset{β}{true}}_{k}

) is known to be biased toward zero if unobserved heterogeneity is present . This bias is caused by the conditional risk set definition of the model, which leads to an imbalance of the unmeasured heterogeneity if a treatment effect is present: consider the case where treatment and some other unmeasured covariate like subjects' age increase the risk of events.…”

Section: Methodsmentioning

confidence: 99%

Safety data from randomized controlled trials: applying models for recurrent events

Hengelbrock

Gillhaus

Kloss

et al. 2016

Pharmaceutical Statistics

View full text Add to dashboard Cite

Simple descriptive listings and inference statistics based on 2×2 tables are still the most common way of summarizing and reporting adverse events data from randomized controlled trials, although these methods do not account for differences in observation times between treatment groups. Using standard methods from survival analysis such as the Cox model or Kaplan-Meier estimates would overcome this problem but limit the analysis to the first safety-related event of each subject. As an alternative, we discuss two models for recurrent events data-the Andersen-Gill and Prentice-Williams-Peterson model-regarding their applicability to safety data from randomized controlled trials. We argue that these models can be used to estimate two different quantities: a direct treatment effect on the risk of an event (Prentice-Williams-Peterson) and a total treatment effect as sum of the direct effect and the treatment's indirect effect via the event history (Anderson-Gill). Using simulated data, we illustrate the difference between these treatment effects and analyze the performance of both models in different scenarios. Because both models are limited to the analysis of cause-specific hazards if competing risks are present, we suggest to incorporate estimates of the mean frequency of events in the analysis to additionally allow the comparison of treatment effects on absolute event probabilities. We demonstrate the application of both models and the mean frequency function to safety endpoints with an illustrative analysis of data from a randomized phase-III study. Copyright © 2016 John Wiley & Sons, Ltd.

show abstract

Empirical study of correlated survival times for recurrent events with proportional hazards margins and the effect of correlation and censoring

Cited by 14 publications

References 13 publications

On the self‐triggering Cox model for recurrent event data

On the self‐triggering Cox model for recurrent event data

Comparison of risks of cardiovascular events in the elderly using standard survival analysis and multiple-events and recurrent-events methods

Safety data from randomized controlled trials: applying models for recurrent events

Contact Info

Product

Resources

About