Treatment Switching: Statistical and Decision-Making Challenges and Approaches

Abstract

Objectives: Treatment switching refers to the situation in a randomized controlled trial where patients switch from their randomly assigned treatment onto an alternative. Often, switching is from the control group onto the experimental treatment. In this instance, a standard intention-to-treat analysis does not identify the true comparative effectiveness of the treatments under investigation. We aim to describe statistical methods for adjusting for treatment switching in a comprehensible way for nonstatisticians, and to summarize views on these methods expressed by stakeholders at the 2014 Adelaide International Workshop on Treatment Switching in Clinical Trials. Methods: We describe three statistical methods used to adjust for treatment switching: marginal structural models, two-stage adjustment, and rank preserving structural failure time models. We draw upon discussion heard at the Adelaide International Workshop to explore the views of stakeholders on the acceptability of these methods. Results: Stakeholders noted that adjustment methods are based on assumptions, the validity of which may often be questionable. There was disagreement on the acceptability of adjustment methods, but consensus that when these are used, they should be justified rigorously. The utility of adjustment methods depends upon the decision being made and the processes used by the decision-maker. Conclusions: Treatment switching makes estimating the true comparative effect of a new treatment challenging. However, many decision-makers have reservations with adjustment methods. These, and how they affect the utility of adjustment methods, require further exploration. Further technical work is required to develop adjustment methods to meet real world needs, to enhance their acceptability to decision-makers.

Bayesian hierarchical meta‐analytic methods for modeling surrogate relationships that vary across treatment classes using aggregate data

“…Patients were allowed to switch from the treatment that was initially assigned to them to the other treatment arm in the trial. Most commonly patients switched after progression from control to experimental arm in particular, if there was sufficient evidence during the trial that the experimental treatment was better than control . Treatment switching has diminishing effect on the difference in treatment effects on OS when applying intention‐to‐treat analysis, and the effect is often obtained with larger uncertainty.…”

Treatment Switching in Cancer Trials: Issues and Proposals

“…These issues are developed in detail in Latimer et al (19). As noted there, there is no single optimal approach to analyzing and interpreting the data from trials where treatment switching has occurred.…”

“…See Latimer et al for a fuller discussion of these issues (19) d. Clinical development programs should consider studies in addition to new trials that can be used to provide more information on: (i) the likely effect of treatment on key trial outcomes, for comparison or combination with data from the trial to allow an estimate of the "true" treatment effects to be made with more confidence; and/or (ii) outcomes not addressed (or inadequately addressed) as a result of omission from the trial protocol, of the powering of the trial, or confounding caused by treatment switching; and/or (iii) the relevance of the trial results for wider patient populations.…”

“…Where adjustment methods are used, trial characteristics and the treatment switching mechanism should be considered alongside the pivotal assumptions associated with each of the statistical adjustment methods to identify which of the analyses (ITT or any of the adjusted analyses) are likely to provide the least amount of bias in each case. Concerns over the validity of methodological assumptions should be addressed as far as possible using the types of analyses described in Latimer et al (19). Detailed justifications should be provided for chosen methods, and consideration should be given to the direction of bias if methodological assumptions are not satisfied.…”

“…In circumstances where the impact of treatment switching is likely to be important, statistical adjustment methods are not the only avenue that may be explored when attempting to address treatment switching. Other options exist (see Latimer et al [19]). These should be discussed with key decision makers as views on their validity and relevance vary.…”