From: Analysis of contamination in cluster randomized trials of malaria interventions
Input: | Geolocations of households; cluster and intervention assignment of households; trial outcome of interest (malaria prevalence or incidence); technical parameters for the MCMC |
Output: | Estimated effectiveness and contamination range with 95%CI |
1: | Calculate distance to the nearest discordant household |
2: | Set up the sigmoid random effects model in JAGS |
3: | Fit the sigmoid random effects model |
4: | Transform the model output for interpretation |