AtlasSession 3L9: Variations in study designp.33–36
Split-part, adaptive, cohort-multiple RCT
- 4 earlier topics
- Normal vs non-normal
- Choosing a statistical test
- Parallel and multi-arm trials
- Cluster and stepped-wedge
- Split-part, adaptive, cohort-multiple RCT
Fig. 1
Sinus grafting (β-TCP granules vs xenograft), a four-scaffold adaptive trial and a 3,000-patient wound cohort (patients)
Fig. 1. 20 people, each with two sides; each side gets one graft, which side is random within the person: 40 units.
- People
- 20
- Sides treated
- 40
- What gets the coin toss
- each side
- Counted as
- linked pairs
Trial Design Board · First look
Level 1 · First look · Lecture 9 · p.33–36
Recognise three unusual trial layouts by what each changes.
Recognise split-part, adaptive and cohort multiple RCT designs and say what each changes: the unit of analysis, the conduct, or recruitment.
Left vs right (the step you are on)
Like two creams on two cheeks, each patient's left and right sinus get different grafts.
Like testing two creams on your two cheeks, each of these 20 patients gets one graft on each side: 40 sides treated, and a coin toss picks which side gets which.