Adjusting survival time estimates to account for treatment switching in randomised controlled trials – a simulation study

Abstract: Background Treatment switching commonly occurs in clinical trials of novel interventions, particularly in the advanced or metastatic cancer setting, which causes important problems for health technology assessment. It is unclear which methods to adjust for switching are most appropriate in realistic scenarios.

Objectives: We aimed to assess statistical approaches for adjusting survival estimates in the presence of treatment switching in order to determine which methods are most appropriate in a range of realistic scenarios.

Methods: We conducted a simulation study to assess the bias, mean squared error and coverage associated with alternative switching adjustment methods across a wide range of realistic scenarios. Results Simple methods such as censoring or excluding patients that switch always resulted in high levels of bias. More complex randomisation-based methods (e.g. Rank Preserving Structural Failure Time Models (RPSFTM)) were unbiased only when the treatment effect was not time-dependent.

Observational-based methods: (e.g. inverse probability of censoring weights (IPCW)) coped better with time-dependent treatment effects but are heavily data reliant, are sensitive to model misspecification and often produced high levels of bias in our simulations, particularly when the proportion of patients that switched treatments was very high (approximately 90%). We introduce a novel “two stage” method, whereby a specific disease-related time-point is used to define a secondary baseline after which switching is permitted, allowing treatment effects to be estimated separately for patients that switch and patients randomised to the experimental group. We find that this method can perform well provided the treatment switching mechanism is amenable.

Conclusions: Randomisation-based methods can accurately adjust for treatment switching when the treatment effect received by patients that switch is the same as that received by patients randomised to the experimental group. When this is not the case observational-based methods or simple twostage methods should be considered, although the IPCW is prone to substantial bias when the proportion of patients that switch is greater than approximately 90%. Simple methods such as censoring or excluding patients that switch should not be used.