Which of the following is the primary mechanism by which randomization reduces bias in a randomized controlled trial?

Explanation

## Randomization and Bias Reduction ### Core Mechanism of Randomization **Key Point:** Randomization distributes both *known* and *unknown* confounders equally between treatment and control groups, thereby isolating the effect of the intervention from confounding variables. ### How Randomization Works 1. **Known confounders** (age, sex, disease severity) are balanced by chance across groups 2. **Unknown confounders** (unmeasured genetic factors, lifestyle variables) are also balanced 3. **Allocation concealment** prevents selection bias by keeping assignment hidden until enrollment 4. **Blinding** (separate from randomization) prevents performance and detection bias ### Why Randomization ≠ Blinding | Feature | Randomization | Blinding | |---------|---------------|----------| | **Prevents** | Selection bias, confounding | Performance bias, detection bias | | **Mechanism** | Random allocation to groups | Concealing group assignment from participants/assessors | | **Can exist alone?** | Yes (open-label RCTs are valid) | No (requires randomization first) | **High-Yield:** Randomization is the *defining feature* of an RCT and the primary reason RCTs are gold-standard evidence. It is NOT the same as blinding, and an RCT can be randomized but unblinded (e.g., surgical intervention trials). **Mnemonic: RACE** — **R**andomization prevents selection bias and confounding; **A**llocation concealment hides assignment; **C**ontrol group provides comparison; **E**vidence of causation emerges. **Clinical Pearl:** Even with randomization, baseline imbalances can occur by chance in small trials. This is why baseline characteristics tables are published — to document that randomization worked. [cite:Park 26e Ch 10]