AtlasSession 4L11: Graphical representation of data and measures of associationp.28–34, 42–47
Risk, odds, RR, OR, RD, NNT
- 1 earlier topics
- Data and types of variables
- Summary statistics
- Summary measures by data type
- Tables
- Risk, odds, RR, OR, RD, NNT
- Next: Other outcome types
Fig. 1
Surgical site infection at 30 days: antibiotic-eluting electrospun mesh vs standard mesh (patients)
Fig. 1. Two groups of patients, one dot each. Antibiotic-eluting: 5 infected out of 50. Standard: 12 infected out of 50. Lit: everyone in each group. Dividing gives 0.42 times; the names for these are risk and risk ratio.
- Antibiotic-eluting: infected
- 5 out of 50
- Standard: infected
- 12 out of 50
- Dividing (risk ratio)
- 0.42 times
- Subtracting (risk difference)
- about 14 fewer in every 100
2×2 Workbench · First look
Level 1 · First look · Lecture 11 · p.28–34, 42–47
Tell 'out of' from 'for every', then compare two arms by dividing and by subtracting.
From a 2×2 table with exposure on the rows, compute risk, odds, RR, OR, RD and NNT/NNH, and explain why the OR drifts from the RR as the control-group risk rises.
Rainy days (the step you are on)
1 wet day out of 7 is the same as 1 wet day for every 6 dry.
One wet day in a week is 1 wet day out of 7, or 1 wet day for every 6 dry. Each well here is one patient: on the eluting mesh 5 were infected and 45 were not.