5. Optimal information size calculation.
| Outcome | Assumed risk | Source | Clinically important relative improvement | Sample size required |
| HIV infection in child | 27/100 | Analysis 1.1 | 25% | 1236 |
| Mean birthweight | 2964 | Analysis 1.2 | 25% | 6178 |
| Low birthweight | 17/100 | Analysis 1.3 | 25% | 2194 |
| Still birth | 3/100 | Analysis 1.4 | 25% | 14,264 |
| Preterm birth | 20/100 | Analysis 1.5 | 25% | 1806 |
| Child death | 14/100 | Analysis 1.6 | 25% | 2748 |
| Maternal death | 3/100 | Analysis 1.7 | 25% | 14,264 |
We based the sample size calculations: 2‐sided tests, with ratio of 1:1, power of 0.8 and confidence level of 0.05. We performed the sample size calculations using http://www.sealedenvelope.com/power/binary‐superiority/