Methods of competing risks analysis of end-stage renal disease and mortality among people with diabetes

Abstract

BackgroundWhen a patient experiences an event other than the one of interest in the study, usually the probability of experiencing the event of interest is altered. By contrast, disease-free survival time analysis by standard methods, such as the Kaplan-Meier method and the standard Cox model, does not distinguish different causes in the presence of competing risks. Alternative approaches use the cumulative incidence estimator by the Cox models on cause-specific and on subdistribution hazards models. We applied cause-specific and subdistribution hazards models to a diabetes dataset with two competing risks (end-stage renal disease (ESRD) or death without ESRD) to measure the relative effects of covariates and cumulative incidence functions.ResultsIn this study, the cumulative incidence curve of the risk of ESRD by the cause-specific hazards model was revealed to be higher than the curves generated by the subdistribution hazards model. However, the cumulative incidence curves of risk of death without ESRD based on those three models were very similar.ConclusionsIn analysis of competing risk data, it is important to present both the results of the event of interest and the results of competing risks. We recommend using either the cause-specific hazards model or the subdistribution hazards model for a dominant risk. However, for a minor risk, we do not recommend the subdistribution hazards model and a cause-specific hazards model is more appropriate. Focusing the interpretation on one or a few causes and ignoring the other causes is always associated with a risk of overlooking important features which may influence our interpretation.

Modelling Length of Hospital Stay for Tuberculosis Treated in-Patients at Queen Elizabeth Central Hospital: a Competing Risk Perspective

“…The results from this study are in agreement with various authors who stated that the SDH model is a better model when studying data that involves competing risks [14,15,26]. Therefore, to meet the objective of identifying prognostic factors of discharge in the presence of competing risks, the sub-distribution hazard model was a better model than the CSH model.…”

“…Competing risk models would be considered the most appropriate for length of stay. There is large body of literature on competing risk models for analysis of time-to-event data in medical research [13][14][15][16]. Despite these studies and papers on LOS, a few studies have looked at modelling length of hospital stay for TB patients and have often not accounted for competing events.…”

“…The cause-specific hazard model may be more clinically understandable when assessing the prognostic effect of the covariates on a specific cause because the covariate effect would be to reduce or increase the instantaneous probability of the event of interest irrespective of other covariate effect. However, when the study objective is to compare the probability of the event of interest, then the sub-distribution hazards model will be appropriate [15]. The sub-distribution model is more desired because it assesses covariate effect on CIF directly unlike cause-specific model which is an indirect measurement.…”

Differential mortality and the excess burden of end-stage renal disease among First Nations people with diabetes mellitus: a competing-risks analysis

“…We previously compared different statistical modelling approaches to optimize the analysis of competing-risks data. 11 Accordingly, for the current study, we used the Fine and Gray model, 12,13 a semiproportional subhazards model that provides the cumulative incidence (or subdistribution) of each event of interest (diagnosis of end-stage renal disease or death before endstage renal disease) while simultaneously considering the competing risk of the other outcome. Thus, people who die before end-stage renal disease occurs are not censored in a way that might bias the estimates, as is possible in a Cox causespecific analysis.…”

“…2,6 We previously evaluated the most appropriate competing-risks methodology for analyzing this kind of data 11 and, on the basis of that evaluation, used Fine and Gray models for the current analysis, as has been proposed by others. 13 The limitations of the study include our inability to control for important predictors of end-stage renal disease and death without end-stage renal disease, such as glycemic, blood pressure and lipid control, and related changes in medical practice, such as the introduction of angiotensinconverting enzyme inhibitors, 30 that occurred during the course of the study period.…”