Comparing Modeling Approaches for Discrete Event Simulations With Competing Risks Based on Censored Individual Patient Data: A Simulation Study and Illustration in Colorectal Cancer

Implementing competing risks in discrete event simulation: the event-specific probabilities and distributions approach

Background: Although several strategies for modelling competing events in discrete event simulation (DES) exist, a methodological gap for the event-specific probabilities and distributions (ESPD) approach when dealing with censored data remains. This study defines and illustrates the ESPD strategy for censored data. Methods: The ESPD approach assumes that events are generated through a two-step process. First, the type of event is selected according to some (unknown) mixture proportions. Next, the times of occurrence of the events are sampled from a corresponding survival distribution. Both of these steps can be modelled based on covariates. Performance was evaluated through a simulation study, considering sample size and levels of censoring. Additionally, an oncology-related case study was conducted to assess the ability to produce realistic results, and to demonstrate its implementation using both frequentist and Bayesian frameworks in R. Results: The simulation study showed good performance of the ESPD approach, with accuracy decreasing as sample sizes decreased and censoring levels increased. The average relative absolute error of the event probability (95%-confidence interval) ranged from 0.04 (0.00; 0.10) to 0.23 (0.01; 0.66) for 60% censoring and sample size 50, showing that increased censoring and decreased sample size resulted in lower accuracy. The approach yielded realistic results in the case study. Discussion: The ESPD approach can be used to model competing events in DES based on censored data. Further research is warranted to compare the approach to other modelling approaches for DES, and to evaluate its usefulness in estimating cumulative event incidences in a broader context.

A Novel Dynamic Approach for Risk Analysis and Simulation Using Multi-Agents Model

Static risk analysis techniques (SRATs) use event graphs and risk analysis assessment models. Those techniques are not time-based techniques and hence are inadequate to model dynamic stochastic systems. This paper proposes a novel dynamic approach to model such stochastic systems using Dynamic Fault Trees (DFT). The proposed model is called Generic Dynamic Agent-Based Model (GDABM) for risk analysis. GDABM is built on top of the well-known Agent-Based Modeling and Simulation (ABMS) technique. GDABM can model the dynamic system agents in both nominal (failure-free) and degraded (failure) modes. GDABM shows the propagation of failure between system elements and provides complete information about the system’s configurations. In this paper, a complete detailed case study is provided to show the GDABM capabilities to model and study the risk analysis for such dynamic systems. In the case study, the GDABM models the risk analysis for a chemical reactor/operator and performs a complete risk analysis for the entire system. The GDABM managed to simulate the dynamic behavior of the system’s components successfully using Repast Simphony 2.0. Detailed agent behavioral modes and failure modes are provided with various scenarios, including different time stamps. The proposed GDABM is compared to a reference model. The reference model is referred to as the ABM model. GDABM has given very promising results. A comparison study was performed on three performance measures. The performance measures used are (1) Accuracy, (2) response time, and (3) execution time. GDABM has outperformed the reference model by 15% in terms of accuracy and by 27% in terms of response time. GDABM incurs a slightly higher execution time (13%) when compared to the ABM reference model. It can be concluded that GDABM can deliver accepted performance in terms of accuracy and response time without incurring much processing overhead.