## Key Principles of RCT Design **High-Yield:** The four pillars of RCT quality are randomization, allocation concealment, blinding, and complete outcome reporting. Selective reporting of only significant outcomes is a form of publication bias and outcome reporting bias — it violates research integrity and is explicitly prohibited in modern trial protocols. ### Why Each Correct Option Is Essential | Principle | Purpose | Bias Prevented | | --- | --- | --- | | **Randomization** | Distribute known and unknown confounders equally between groups | Selection bias, confounding | | **Allocation Concealment** | Prevent investigators from knowing treatment assignment in advance | Selection bias | | **Blinding** | Mask treatment assignment from participants and assessors | Performance bias, detection bias | | **Complete Outcome Reporting** | Report all pre-specified outcomes, regardless of direction or magnitude | Outcome reporting bias, publication bias | ### Why Selective Reporting Is Wrong **Key Point:** Selective reporting of only statistically significant outcomes is a form of **outcome reporting bias** and **publication bias**. It: 1. Inflates effect sizes and false-positive rates 2. Distorts the evidence base and misleads clinical decision-making 3. Violates the CONSORT checklist and trial registration requirements 4. Is considered research misconduct in modern guidelines **Clinical Pearl:** The CONSORT 2010 statement and the EQUATOR network explicitly mandate pre-registration of all outcomes and complete reporting of results, whether positive, negative, or null. **Warning:** Confuse this with "post-hoc analysis" — which is permissible if labeled as exploratory, not confirmatory. The trap here is selective *reporting* of pre-specified outcomes, which is never acceptable.
Sign up free to access AI-powered MCQ practice with detailed explanations and adaptive learning.