Related Experiment Videos
Multistage test procedure for testing Blackwelder's hypothesis of equivalence
M Y Lakshminarayanan1, H I Patel, W J Stager
1Pfizer, Inc., Groton, Connecticut 06340.
Journal of Biopharmaceutical Statistics
|July 1, 1994
Summary
This study modifies group sequential procedures for clinical trials to assess if a new therapy is not worse than a standard one by a specific margin. The findings are applicable to drug interaction studies focusing on side effects.
Area of Science:
- Clinical Trials Methodology
- Biostatistics
- Pharmacology
Background:
- Assessing non-inferiority in clinical trials is crucial for new therapy adoption.
- Existing group sequential methods by DeMets and Ware are limited to delta = 0 (no difference).
- Blackwelder's hypothesis addresses showing experimental therapy is not worse than standard by a margin (delta).
Purpose of the Study:
- To modify DeMets and Ware's asymmetric group sequential procedure for non-inferiority trials where delta > 0.
- To adapt the procedure for a two-stage design in drug interaction studies.
- To evaluate experimental therapy effectiveness against a standard therapy with a defined margin of equivalence.
Main Methods:
- Modification of the asymmetric group sequential procedure by DeMets and Ware.
- Application of a two-stage design for interim and final analyses.
- Focus on dichotomous outcomes, specifically a side effect in drug interaction studies.
Main Results:
- The modified procedure accommodates non-inferiority testing with delta > 0.
- A two-stage design was successfully applied to a drug interaction study scenario.
- The methodology provides a framework for analyzing superiority/non-inferiority with a specified margin.
Conclusions:
- The modified group sequential procedure effectively handles non-inferiority testing for delta > 0.
- This approach is valuable for clinical trial design, particularly in drug interaction studies.
- The study offers a statistically sound method for comparing therapies with a non-inferiority margin.