Treatment 2×2 — RR, OR, NNT
§2.9
Rev 4.6 · reviewed Feb 2026Treatment 2×2 — RR, OR, NNT
For a binary outcome under two treatments. Enter event and no-event counts for each arm.
Source formula: Altman DG. BMJ 1998;317:1309–12.
Table. Events by treatment arm
| Event | No event | Total | Risk | |
|---|---|---|---|---|
| Treated | 150 | 8.0% | ||
| Control | 150 | 16.0% |
* Ratio CIs computed on the log scale.
† NNT interval per Altman's benefit → ∞ → harm convention when the ARR CI crosses zero. Altman DG. BMJ 1998;317:1309–12.
Risk ratio (95% CI)*
0.50 (0.26–0.96)▲
copyOdds ratio (95% CI)*
0.46 (0.22–0.95)
copyAbsolute risk reduction (95% CI)
8.0% (0.7% to 15.3%)▲
copyNNT (95% CI)†
12.5 · 6.5 to 142.6
copyTreatment effect
RR 0.5095% CI 0.26–0.96
Absolute reduction 8.0%; about 13 patients treated to prevent one event.
Treating 13 patients prevents about one event: risk falls from 16.0% to 8.0%. Both the relative effect (RR 0.50) and the absolute effect matter — report both.
Understand what this is measuring