6. Sample size suggestions for trial of antibiotics in Shigella dysentery.

SAMPLE SIZES	Antibiotic versus no drug or placebo (placebo response at 45%) or Antibiotic versus another antibiotic
1 sided α	10% difference: 310 20% difference: 75 25% difference: 50 30% difference: 30 40% difference: 15
2 sided α	10% difference: 390 20% difference: 95 25% difference: 60 30% difference: 40 40% difference: 20

1. The sample size required to detect the assumed difference in improvement or worsening with 80% power and 5% significance level using the outcome of 'diarrhoea at follow up' from this review using StatCalc 2006.
 2. The sample size mentioned is for each arm of the study.