Regarding the validity and quality of randomized controlled trials, all of the following statements are TRUE EXCEPT:

Explanation

## Internal and External Validity in RCTs ### Defining the Two Pillars of RCT Quality **Key Point:** Internal validity (freedom from bias within the study) and external validity (applicability to broader populations) are distinct concepts that often exist in tension. ### Internal Validity — Minimized by Randomization & Blinding **High-Yield:** Internal validity is threatened by: - **Selection bias** — non-random allocation or differential enrollment - **Performance bias** — differential treatment or care between groups - **Detection bias** — differential outcome assessment - **Attrition bias** — differential loss to follow-up All are effectively minimized by proper randomization, allocation concealment, and blinding. ✓ **Option 0 is CORRECT.** ### External Validity — REDUCED (Not Enhanced) by Strict Selection Criteria **Warning:** This is a common misconception. Strict inclusion/exclusion criteria improve **internal validity** (by reducing confounding and heterogeneity) but **reduce external validity** (generalizability). **Clinical Pearl:** A highly selected, homogeneous population: - ✓ Reduces confounding and noise → stronger internal validity - ✗ Limits applicability to real-world, diverse populations → weaker external validity Example: An RCT enrolling only 40–50-year-old men with uncomplicated hypertension and no comorbidities has high internal validity but poor generalizability to elderly patients, women, or those with diabetes or CKD. **Mnemonic:** **STRICT = Strong Internal, Reduced Transfer** (to external populations) ### Intention-to-Treat (ITT) Analysis **Key Point:** ITT analysis: - Preserves the benefits of randomization (unbiased group comparison) - Includes all randomized participants in their assigned group, regardless of adherence or completion - Provides a **pragmatic, conservative** estimate of treatment effect - Is the gold standard for efficacy trials and is mandated by regulatory agencies (FDA, EMA) ✓ **Option 2 is CORRECT.** ### Attrition Bias & Loss to Follow-up **High-Yield:** Loss to follow-up is a major threat to internal validity: - **>20% loss** in either arm is considered substantial and high-risk for bias - **>10% differential loss** between arms is particularly concerning - Must be reported transparently (CONSORT checklist item 17a) - Sensitivity analyses should be conducted to assess impact ✓ **Option 3 is CORRECT.** ### Summary Table: Internal vs. External Validity | Factor | Effect on Internal Validity | Effect on External Validity | |--------|-------|----------| | **Strict inclusion/exclusion criteria** | ↑ (reduces confounding) | ↓ (limits generalizability) | | **Randomization & blinding** | ↑↑ (prevents bias) | ↔ (neutral) | | **Homogeneous population** | ↑ (reduces noise) | ↓ (poor representation) | | **Diverse, pragmatic enrollment** | ↓ (more confounding) | ↑ (better generalizability) | **Tip:** When appraising an RCT, ask: *"Does this trial answer the efficacy question (internal validity) or the effectiveness question (external validity)?" Often, you need both — efficacy from a tightly controlled RCT, then effectiveness from pragmatic trials or observational studies in real-world populations.*