Related Experiment Videos
Hazards of inference: the active control investigation
1Food and Drug Administration, Rockville, Maryland.
Epilepsia
|January 1, 1989
Summary
Finding no difference between epilepsy drugs doesn't prove new drug efficacy. Other factors, not just drug action, can explain similar results in clinical trials.
Area of Science:
- Neurology
- Clinical Pharmacology
- Biostatistics
Background:
- Active control studies are popular for evaluating new antiepileptic drugs.
- A null difference in seizure outcomes is often misinterpreted as proof of efficacy.
Purpose of the Study:
- To critically evaluate the interpretation of "no significant difference" findings in antiepileptic drug trials.
- To highlight the limitations of active control studies in demonstrating anticonvulsant efficacy.
Main Methods:
- Analysis of the logical and evidential basis for inferring drug efficacy from null results.
- Review of common assumptions and potential confounding factors in clinical trial design.
Main Results:
- A finding of no difference between an experimental and standard drug does not inherently confirm the experimental drug's efficacy.
- Numerous factors unrelated to pharmacological action can lead to null results in comparative studies.
Conclusions:
- The interpretation of null differences requires external evidence beyond the study itself.
- Relying on a series of non-significant findings to infer efficacy carries substantial risk due to study design deficiencies.