Related Experiment Videos
Stopping guidelines for clinical trials with multiple treatments
1Department of Biostatistics, Harvard School of Public Health, Boston, MA 02115.
Statistics in Medicine
|May 30, 1993
Summary
New stopping guidelines allow multiple treatments in clinical trials to be dropped if they are inferior. This method preserves the statistical error rate for identifying the best treatment, even with standard of care comparisons.
Area of Science:
- Clinical Trials
- Biostatistics
- Medical Research
Background:
- Stopping guidelines are crucial in long-term clinical trials with two treatments to maintain statistical integrity during interim analyses.
- Extending these guidelines to multiple treatments allows for the ethical and efficient discontinuation of underperforming therapies.
Purpose of the Study:
- To adapt and validate stopping guidelines for multi-arm clinical trials, preserving the type I error rate.
- To ensure accurate identification of the best treatment or group of treatments in comparative studies.
Main Methods:
- Simulation studies were used to evaluate the performance of stopping guidelines in multi-arm trials.
- The O'Brien and Fleming method for stopping boundaries was applied to normally distributed responses in trials with 3-4 treatments.
Main Results:
- Stopping guidelines developed for two-arm trials are directly applicable to multi-arm trials with only experimental treatments.
- A modified guideline allows for dropping treatments inferior to a standard of care, enhancing ethical considerations.
Conclusions:
- The proposed stopping guidelines are effective for multi-arm trials, preserving statistical error rates.
- Considerations for implementing these guidelines in complex multi-treatment trial designs are discussed.