A researcher is designing a randomized controlled trial to evaluate the efficacy of a new antihypertensive drug in Indian adults. To minimize bias and ensure valid comparison between treatment and control groups, which investigation should be performed at baseline to assess the comparability of groups?

Explanation

## Baseline Assessment in RCT Design **Key Point:** Baseline investigations in an RCT serve to document the comparability of randomized groups and establish the starting point for outcome measurement. They are NOT interventions but rather descriptive assessments. ### Purpose of Baseline Investigations Baseline data collection in RCTs achieves two critical objectives: 1. **Verification of group comparability** — ensures randomization was effective 2. **Adjustment for confounders** — allows stratified analysis if imbalances occur ### What Should Be Documented at Baseline | Element | Rationale | |---------|----------| | Demographic variables (age, sex, occupation) | Establish population characteristics | | Clinical baseline measurements (BP, weight, comorbidities) | Quantify starting disease severity | | Relevant biomarkers (creatinine, lipids if applicable) | Document organ function status | | Inclusion/exclusion criteria verification | Confirm eligibility | | Randomization method documentation | Prove allocation concealment | **High-Yield:** The gold standard is **stratified randomization** with **complete baseline documentation**. This ensures: - Balanced distribution of known confounders - Ability to perform subgroup analyses - Transparency in reporting (CONSORT guidelines) ### Why Baseline Investigations Are Universal **Clinical Pearl:** Even in double-blind RCTs, baseline investigations are performed openly because they establish the reference point for measuring change. They are not blinded because they precede randomization. **Mnemonic: BASELINE** — **B**iomarkers, **A**ge/demographics, **S**everity (clinical), **E**ligibility confirmation, **L**aboratory values, **I**nclusion criteria, **N**eed for stratification, **E**quivalence verification. [cite:Park 26e Ch 10]