Use and reporting of restricted randomization: a review.

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Research Integrity

Background:

• Restricted randomization (blocking, minimization) balances groups but risks bias.
• Current usage and reporting of these methods are not well-documented.
• Inadequate reporting hinders assessment of bias risk.

Purpose of the Study:

• Assess the frequency of restricted randomization methods in clinical trials.
• Evaluate the reporting of these methods against established standards (e.g., CONSORT).
• Identify potential risks of selection bias and technical error due to randomization practices.

Main Methods:

• Reviewed 82 reports of randomized controlled trials.
• Recorded and assessed the reported randomization methods.
• Consulted protocols, other materials, and authors for clarification when needed.

Main Results:

• 88% of trials used some form of restricted randomization.
• Randomization methods were not reported or lacked detail in 50% of trials.
• Predictable methods (small blocks, minimization without randomness) were used in 19% of trials.

Conclusions:

• The majority of trials employ restricted randomization, often with predictable methods.
• Inadequate reporting of randomization compromises the ability to assess bias risk.
• Improved reporting is crucial for ensuring the integrity of randomized controlled trials.