Semicompeting risks in aging research: methods, issues and needs

Abstract

A semicompeting risks problem involves two-types of events: a nonterminal and a terminal event (death). Typically, the nonterminal event is the focus of the study, but the terminal event can preclude the occurrence of the nonterminal event. Semicompeting risks are ubiquitous in studies of aging. Examples of semicompeting risk dyads include: dementia and death, frailty syndrome and death, disability and death, and nursing home placement and death. Semicompeting risk models can be divided into two broad classes: models based only on observables quantities (class scriptO) and those based on potential (latent) failure times (class scriptL). The classical illness-death model belongs to class scriptO. This model is a special case of the multistate models, which has been an active area of methodology development. During the past decade and a half, there has also been a flurry of methodological activity on semicompeting risks based on latent failure times (scriptL models). These advances notwithstanding, the semi-competing risks methodology has not penetrated biomedical research, in general, and gerontological research, in particular. Some possible reasons for this lack of uptake are: the methods are relatively new and sophisticated, conceptual problems associated with potential failure time models are difficult to overcome, paucity of expository articles aimed at educating practitioners, and non-availability of readily usable software. The main goals of this review article are: (i) to describe the major types of semicompeting risks problems arising in aging research, (ii) to provide a brief survey of the semicompeting risks methods, (iii) to suggest appropriate methods for addressing the problems in aging research, (iv) to highlight areas where more work is needed, and (v) to suggest ways to facilitate the uptake of the semicompeting risks methodology by the broader biomedical research community.

Biased Exposure–Health Effect Estimates from Selection in Cohort Studies: Are Environmental Studies at Particular Risk?

“…Because our outcome was mortality, we ignored selection due to death and estimated our effect among the survivors at the time of the KXRF. Although IPW and other methods can be used to address selection due to death, the utility of such efforts is controversial ( Chaix et al 2012 ; Tchetgen Tchetgen et al 2012 ; Weuve et al 2012 ), and we refer the reader elsewhere ( Andersen and Keiding 2012 ; Chaix et al 2012 ; Kurland et al 2009 ; Lau et al 2009 ; Tchetgen Tchetgen 2014 ; Tchetgen Tchetgen et al 2012 ; Varadhan et al 2014 ; Weuve et al 2012 ).…”

“…IPW uses information available for participants with and without KXRF measurements to weight observations from participants with a KXRF measurement, so that the weighted subpopulation is representative of all NAS participants who are alive at the time of the KXRF substudy. [It is possible to address attrition by death with this technique as well; however, this is controversial and particularly problematic for the current study, where the outcome is mortality—we refer readers to the work of others for further consideration of what to do with attrition due to death ( Andersen and Keiding 2012 ; Chaix et al 2012 ; Kurland et al 2009 ; Lau et al 2009 ; Tchetgen Tchetgen 2014 ; Tchetgen Tchetgen et al 2012 ; Varadhan et al 2014 ; Weuve et al 2012 )]. In this way the arrows into S KXRF are removed (because the group is weighted to be representative of the whole living NAS population and, therefore, not a selected group of the original NAS study sample), thus eliminating the bias induced by conditioning on a collider or conditioning on the intermediate CV 1 through conditioning on its descendant S KXRF ( Figure 5D ).…”

Natural history of diseases: Statistical designs and issues

“…There is a rich literature on these topics. 45 The study by Varadhan et al 46 provide an introduction to some of the relevant ideas and methods.…”

SemiCompRisks: An R Package for the Analysis of Independent and Cluster-correlated Semi-competing Risks Data

“…The current literature for the analysis of semi-competing risks data is composed of three approaches: methods that specify the dependence between non-terminal and terminal events via a copula (Fine et al, 2001;Wang, 2003;Jiang et al, 2005;Ghosh, 2006;Peng and Fine, 2007;Lakhal et al, 2008;Hsieh et al, 2008;Fu et al, 2013); methods based on multi-state models, specifically the so-called illness-death model (Liu et al, 2004;Putter et al, 2007;Ye et al, 2007;Kneib and Hennerfeind, 2008;Zeng and Lin, 2009;Xu et al, 2010;Zeng et al, 2012;Han et al, 2014;Zhang et al, 2014;Lee et al, 2015; and methods built upon the principles of causal inference (Zhang and Rubin, 2003;Egleston et al, 2007;Tchetgen Tchetgen, 2014;Varadhan et al, 2014).…”