A community randomized trial (cluster RCT) differs from an individual RCT primarily in that:
- A Randomization is of groups, producing clustering of outcomes that inflates type I error if not accounted for in analysis ✓
- B Randomization is of groups (villages/schools) rather than individuals, increasing statistical power
- C It is always superior to individual RCT in controlling confounding
- D Blinding is easier in cluster RCTs
Correct answer: A. Randomization is of groups, producing clustering of outcomes that inflates type I error if not accounted for in analysis
Explanation
In cluster RCTs, entire groups (clusters) are randomized rather than individuals. Individuals within a cluster are more similar to each other than to those in other clusters (intraclass correlation, ICC > 0), so outcomes are not independent — this is 'clustering effect.' If clustering is ignored in analysis, standard errors are underestimated and type I errors are inflated. The design effect = 1 + (m-1)×ICC must be applied; sample sizes must be inflated accordingly.
Reference: Park's Textbook of Preventive and Social Medicine, 27th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.