Which of the following is the primary purpose of randomization in a randomized controlled trial?

Explanation

## Purpose of Randomization in RCTs **Key Point:** Randomization is the cornerstone of RCT design. Its primary purpose is to distribute both **known and unknown confounders** equally between groups, thereby eliminating confounding bias and establishing causal inference. ### How Randomization Achieves This 1. **Balances known confounders** — age, sex, disease severity, comorbidities are distributed evenly by chance 2. **Balances unknown confounders** — variables not yet identified or measured are also balanced 3. **Enables causal inference** — differences in outcome between groups can be attributed to the intervention, not to baseline differences ### Key Mechanisms **High-Yield:** Randomization is the ONLY method that balances unknown confounders. Matching, stratification, and statistical adjustment can only address known variables. **Mnemonic:** **R-A-N-D-O-M = Removes All Non-Differential Observation Mixing** — randomization eliminates systematic differences in group composition. ### What Randomization Does NOT Guarantee | Misconception | Reality | |---------------|----------| | Equal doses | Randomization assigns groups, not dosing schedules | | Statistical significance | Randomization enables causal inference; significance depends on effect size and sample size | | Smaller sample size | Randomization does not reduce sample size requirements | | Perfect balance | Random imbalance can occur in small samples (use stratified randomization if needed) | **Clinical Pearl:** Even with perfect randomization, baseline imbalances can occur by chance, especially in small trials. This is why baseline characteristics tables are reported — to document that randomization worked. **Warning:** Do not confuse randomization (assignment to groups) with blinding (masking group assignment from participants/assessors). Both are important but serve different purposes. [cite:Park 26e Ch 8]