Research protocol of two concurrent cluster-randomized trials: Real-life Effect of a CAMPaign with Measles Vaccination (RECAMP-MV) and Real-life Effect of a CAMPaign with Oral Polio Vaccination (RECAMP-OPV) on mortality and morbidity among children in rural Guinea-Bissau

Table 2 Sample size estimates derived from the first complete round of enrolment and follow-up visit of RECAMP-MV and RECAMP-OPV

RECAMP	MV	OPV
Number of clusters	222
Alpha	0.05
Between cluster variation coefficient	0.25
Number of eligible children to be enrolled	18,000	10,000
Observed non-accidental deaths/non-accidental hospitalizations rates	15/1000 pyrs*	48/1000 pyrs**
Harmonic mean of total projected accumulated observation time per cluster	84 pyrs	40 pyrs
Expected reduction	30%	25%
Power	80%	80%

*Assuming that our observed composite outcome rate (15/1000 pyrs) is an average of the rates in our control and intervention clusters, and that the real difference between the clusters is 30%, we assumed the rates to be 17/1000 pyrs in control clusters and 12/1000 pyrs in intervention clusters when re-evaluating our power calculations
**Assuming that our observed composite outcome rate (48/1000 pyrs) is an average of the rates in our control and intervention clusters, and that the real difference between the clusters is 25%, we assumed the rates to be 55/1000 pyrs in control clusters and 41/1000 pyrs in intervention clusters when re-evaluating our power calculations
Abbreviation: Pyrs = person-years at risk

ISSN: 1471-2458