Methods for Analyzing Treatment Switching in Clinical Trials
Author Information
Author(s): James P. Morden, Paul C. Lambert, Nicholas Latimer, Keith R. Abrams, Allan J. Wailoo
Primary Institution: The Institute of Cancer Research
Hypothesis
What is the best method to analyze clinical trials with treatment switching?
Conclusion
Simple methods are often not appropriate to deal with treatment switching, and alternative approaches like the Branson & Whitehead method should be considered.
Supporting Evidence
- The study simulated 1000 datasets for each of 16 different scenarios.
- Selection bias was observed in simple methods when survival differences were large.
- The Branson & Whitehead method provided less biased estimates in scenarios with treatment switching.
Takeaway
This study looks at how to analyze clinical trials when patients switch treatments, finding that some methods can give misleading results.
Methodology
A simulation study was conducted to assess the performance of various methods for analyzing treatment switching in clinical trials.
Potential Biases
Some methods may introduce selection bias if patients who switch are not representative of the overall trial population.
Limitations
The study only considered scenarios where patients switch from control to experimental treatment and did not explore switching in both directions.
Participant Demographics
Patients were split into 'good' and 'poor' prognosis groups.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website