Related Experiment Videos
Equivalence testing with dental clinical trials
1Department of Prosthodontics, School of Dentistry, Virginia Commonwealth University, Richmond 23298-0566, USA. Dburns@den1.den.vcu.edu
Journal of Dental Research
|August 14, 2001
Summary
Equivalence testing directly demonstrates if dental treatments are comparable, unlike traditional superiority tests. This method, particularly crossover designs, is efficient for dental clinical trials, requiring smaller sample sizes than parallel designs.
Area of Science:
- Dental Research
- Clinical Trials
- Biostatistics
Background:
- Traditional clinical trials often infer treatment equivalence from a lack of superiority.
- Equivalence testing methods, well-established in pharmaceuticals for bioequivalence, are underutilized in dental research.
- Existing dental literature rarely documents the application of direct equivalence testing.
Purpose of the Study:
- To introduce and discuss the application of equivalence testing in dental clinical trials.
- To contrast equivalence testing with traditional superiority testing.
- To explore the suitability of parallel and crossover designs for dental equivalence trials.
Main Methods:
- Contrasting equivalence and superiority testing methodologies.
- Comparing parallel and crossover designs within the context of equivalence testing.
- Illustrating equivalence trial methodology with a dental implant overdenture example.
Main Results:
- Equivalence testing offers a direct approach to demonstrating treatment comparability.
- Crossover designs are significantly more sample-efficient for equivalence trials than parallel designs.
- For 90% power, a crossover trial requires 24 subjects, while a parallel design needs over 150.
Conclusions:
- Equivalence testing, especially using crossover designs, is an efficient and appropriate methodology for dental clinical trials.
- The adoption of equivalence testing can enhance the rigor and efficiency of dental research.
- Further application of these statistical methods in dental research is recommended.