• Journal of Applied Reliability
• /
• v.18 no.2
• /
• pp.130-142
• /
• 2018

Purpose: The purpose of this study is to introduce regression method in the presence of competing risks and to show how you can use the method with hypothetical data. Methods: Survival analysis has been widely used in biostatistics division. But the same method has not been utilized in reliability division. Especially competing risks, where more than a couple of causes of failure occur and the occurrence of one event precludes the occurrence of the other events, are scattered in reliability field. But they are not utilized in the area of reliability or they are analysed in the wrong way. Specifically Kaplan-Meier method is used to calculate the probability of failure in the presence of competing risks, thereby overestimating the real probability of failure. Hence, cumulative incidence function is introduced. In addition, sample competing risks data are analysed using cumulative incidence function along with some graphs. Lastly we compare cumulative incidence functions with regression type analysis briefly. Results: We used cumulative incidence function to calculate the survival probability or failure probability in the presence of competing risks. We also drew some useful graphs depicting the failure trend over the lifetime. Conclusion: This research shows that Kaplan-Meier method is not appropriate for the evaluation of survival or failure over the course of lifetime in the presence of competing risks. Cumulative incidence function is shown to be useful in stead. Some graphs using the cumulative incidence functions are also shown to be informative.

• Communications for Statistical Applications and Methods
• /
• v.25 no.4
• /
• pp.385-396
• /
• 2018

Competing risks are commonly encountered in biomedical research. Regression models for competing risks data can be developed based on data routinely collected in hospitals or general practices. However, these data sets usually contain the covariate missing values. To overcome this problem, multiple imputation is often used to fit regression models under a MAR assumption. Here, we introduce a multivariate imputation in a chained equations algorithm to deal with competing risks survival data. Using pseudo-observations, we make use of the available outcome information by accommodating the competing risk structure. Lastly, we illustrate the practical advantages of our approach using simulations and two data examples from a coronary artery disease data and hepatocellular carcinoma data.

• The Korean Journal of Applied Statistics
• /
• v.29 no.7
• /
• pp.1231-1246
• /
• 2016

Cause-specific hazard model (Prentice et al., 1978) and subdistribution hazard model (Fine and Gray, 1999) are mostly used for the right censored survival data with competing risks. Some other models for survival data with competing risks have been subsequently introduced; however, those models have not been popularly used because the models cannot provide reliable statistical estimation methods or those are overly difficult to compute. We introduce simple and reliable competing risk regression models which have been recently proposed as well as compare their methodologies. We show how to use SAS and R for the data with competing risks. In addition, we analyze survival data with two competing risks using five different models.

• Journal of Applied Reliability
• /
• v.16 no.1
• /
• pp.56-63
• /
• 2016

Purpose: The purpose of this study is to point out that the Kaplan-Meier method is not valid to calculate the survival probability or failure probability (risk) in the presence of competing risks and to introduce more valid method of cumulative incidence function. Methods: Survival analysis methods have been widely used in biostatistics division. However the same methods have not been utilized in reliability division. Especially competing risks cases, where several causes of failure occur and the occurrence of one event precludes the occurrence of the other events, are scattered in reliability field. But they are not noticed in the realm of reliability expertism or they are analysed in the wrong way. Specifically Kaplan-Meier method which assumes that the censoring times and failure times are independent is used to calculate the probability of failure in the presence of competing risks, thereby overestimating the real probability of failure. Hence, cumulative incidence function is introduced and sample competing risks data are analysed using cumulative incidence function and some graphs. Finally comparison of cumulative incidence functions and regression type analysis are mentioned briefly. Results: Cumulative incidence function is used to calculate the survival probability or failure probability (risk) in the presence of competing risks and some useful graphs depicting the failure trend over the lifetime are introduced. Conclusion: This paper shows that Kaplan-Meier method is not appropriate for the evaluation of survival or failure over the course of lifetime. In stead, cumulative incidence function is shown to be useful. Some graphs using the cumulative incidence functions are also shown to be informative.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.555-562
• /
• 2016

Interval censored data often occur in an observational study where the subject is followed periodically. Instead of observing an exact failure time, two inspection times that include it are available. There are several methods to analyze interval censored failure time data (Sun, 2006). However, in the presence of competing risks, few methods have been suggested to estimate covariate effect on interval censored competing risk data. A sub-distribution hazard model is a commonly used regression model because it has one-to-one correspondence with a cumulative incidence function. Alternatively, Klein and Andersen (2005) proposed a pseudo-value approach that directly uses the cumulative incidence function. In this paper, we consider an extension of the pseudo-value approach into the interval censored data to estimate regression coefficients. The pseudo-values generated from the estimated cumulative incidence function then become response variables in a generalized estimating equation. Simulation studies show that the suggested method performs well in several situations and an HIV-AIDS cohort study is analyzed as a real data example.

• The Korean Journal of Applied Statistics
• /
• v.29 no.7
• /
• pp.1311-1327
• /
• 2016

We propose a multi-state model for analyzing semi-competing risks data with interval-censored or missing intermediate events. This model is an extension of the 'illness-death model', which composes three states, such as 'healthy', 'diseased', and 'dead'. The state of 'diseased' can be considered as an intermediate event. Two more states are added into the illness-death model to describe missing events caused by a loss of follow-up before the end of the study. One of them is a state of 'LTF', representing a lost-to-follow-up, and the other is an unobservable state that represents the intermediate event experienced after LTF occurred. Given covariates, we employ the Cox proportional hazards model with a normal frailty and construct a full likelihood to estimate transition intensities between states in the multi-state model. Marginalization of the full likelihood is completed using the adaptive Gaussian quadrature, and the optimal solution of the regression parameters is achieved through the iterative Newton-Raphson algorithm. Simulation studies are carried out to investigate the finite-sample performance of the proposed estimation procedure in terms of the empirical coverage probability of the true regression parameter. Our proposed method is also illustrated with the dataset adapted from Helmer et al. (2001).

• The Korean Journal of Applied Statistics
• /
• v.33 no.4
• /
• pp.379-393
• /
• 2020

A terminal event such as death may censor an intermediate event such as relapse, but not vice versa in semi-competing risks data, which is often seen in medicine, public health, and epidemiology. We propose a Weibull regression model with a normal frailty to analyze semi-competing risks data when all three transition times of the illness-death model are possibly interval-censored. We construct the conditional likelihood separately depending on the types of subjects: still alive with or without the intermediate event, dead with or without the intermediate event, and dead with the intermediate event missing. Optimal parameter estimates are obtained from the iterative quasi-Newton algorithm after the marginalization of the full likelihood using the adaptive importance sampling. We illustrate the proposed method with extensive simulation studies and PAQUID (Personnes Agées Quid) data.

• The Korean Journal of Applied Statistics
• /
• v.30 no.4
• /
• pp.539-553
• /
• 2017

We propose a multi-state model to analyze semi-competing risks data with interval-censored or missing intermediate events. This model is an extension of the three states of the illness-death model: healthy, disease, and dead. The 'diseased' state can be considered as the intermediate event. Two more states are added into the illness-death model to incorporate the missing events, which are caused by a loss of follow-up before the end of a study. One of them is a state of the lost-to-follow-up (LTF), and the other is an unobservable state that represents an intermediate event experienced after the occurrence of LTF. Given covariates, we employ the Lin and Ying additive hazards model with log-normal frailty and construct a conditional likelihood to estimate transition intensities between states in the multi-state model. A marginalization of the full likelihood is completed using adaptive importance sampling, and the optimal solution of the regression parameters is achieved through an iterative quasi-Newton algorithm. Simulation studies are performed to investigate the finite-sample performance of the proposed estimation method in terms of empirical coverage probability of true regression parameters. Our proposed method is also illustrated with a dataset adapted from Helmer et al. (2001).

• The Korean Journal of Applied Statistics
• /
• v.37 no.2
• /
• pp.225-237
• /
• 2024

In the analysis of competing risks data, some of covariates may not be fully observed for some subjects. In such cases, excluding subjects with missing covariate values from the analysis may result in biased estimates and loss of efficiency. In this paper, we studied multiple imputation and the augmented inverse probability weighting method for regression parameter estimation in the cause-specific proportional hazards model with missing covariates. The performance of estimators obtained from multiple imputation and the augmented inverse probability weighting method is evaluated by simulation studies, which show that those methods perform well. Multiple imputation and the augmented inverse probability weighting method were applied to investigate significant risk factors for the risk of death from breast cancer and from other causes for breast cancer data with missing values for tumor size obtained from the Prostate, Lung, Colorectal, and Ovarian Cancer Screen Trial Study. Under the cause-specific proportional hazards model, the methods show that race, marital status, stage, grade, and tumor size are significant risk factors for breast cancer mortality, and stage has the greatest effect on increasing the risk of breast cancer death. Age at diagnosis and tumor size have significant effects on increasing the risk of other-cause death.

• Asian Pacific Journal of Cancer Prevention
• /
• v.14 no.9
• /
• pp.5355-5360
• /
• 2013

Background: Clarifying the prognostic impact of histological type is an essential issue that may influence the treatment and follow-up planning of newly diagnosed cervical cancer cases. This study aimed to evaluate the prognostic impact of histological type on survival and mortality in patients with cervical squamous cell carcinoma (SCC), adenocarcinoma (ADC) and small cell neuroendocrine carcinoma (SNEC). Materials and Methods: All patients with cervical cancer diagnosed and treated at Chiang Mai University Hospital between January 1995 and October 2011 were eligible. We included all patients with SNEC and a random weighted sample of patients with SCC and ADC. We used competing-risks regression analysis to evaluate the association between histological type and cancer-specific survival and mortality. Results: Of all 2,108 patients, 1,632 (77.4%) had SCC, 346 (16.4%) had ADC and 130 (6.2%) had SNEC. Overall, five-year cancer-specific survival was 60.0%, 54.7%, and 48.4% in patients with SCC, ADC and SNEC, respectively. After adjusting for other clinical and pathological factors, patients with SNEC and ADC had higher risk of cancer-related death compared with SCC patients (hazard ratio [HR] 2.6; 95% CI, 1.9-3.5 and HR 1.3; 95% CI, 1.1-1.5, respectively). Patients with SNEC were younger and had higher risk of cancer-related death in both early and advanced stages compared with SCC patients (HR 4.9; 95% CI, 2.7-9.1 and HR 2.5; 95% CI, 1.7-3.5, respectively). Those with advanced-stage ADC had a greater risk of cancer-related death (HR 1.4; 95% CI, 1.2-1.7) compared with those with advanced-stage SCC, while no significant difference was observed in patients with early stage lesions. Conclusion: Histological type is an important prognostic factor among patients with cervical cancer in Thailand. Though patients with SNEC were younger and more often had a diagnosis of early stage compared with ADC and SCC, SNEC was associated with poorest survival. ADC was associated with poorer survival compared with SCC in advanced stages, while no difference was observed at early stages. Further tailored treatment-strategies and follow-up planning among patients with different histological types should be considered.