Sample size of randomized double-blind trials 1976-1991

Area of Science:

• Clinical Research
• Biostatistics
• Evidence-Based Medicine

Background:

• The adequacy of sample sizes in clinical trials is crucial for reliable results.
• Historical trends in randomized trial sample sizes are not well-documented.

Purpose of the Study:

• To investigate the trend in sample size of randomized trials over recent decades.
• To assess if sample sizes have increased and if they are sufficient.

Main Methods:

• Systematic review of randomized trials published in 1976, 1981, 1986, and 1991.
• Inclusion criteria: double-blind, active treatments, non-crossover, English full articles, clinical outcomes.

Main Results:

• Median sample size increased from 46 (1976) to 71 (1991).
• Specialties like gastroenterology, cardiology, and oncology had larger sample sizes.
• Use of binary outcomes and trials with significant results increased over time.

Conclusions:

• Despite an increase, most randomized trial sample sizes are still insufficient.
• Hundreds of patients are needed to reliably detect a 25% treatment improvement.
• Regulatory bodies should reject small trials unless for rare diseases.