Thanks Lisa, this is an improtant point. I think the "herd immunity" is an important strength of the intervention. It also presents challenges for our evaluation. Due to sample size issues and the limited geographic scope of our study area, we are conducting randomizing the intervention at the individual patient level. This may lead to cross-over between study arms, or spill-over effect of the intervention. Something we will try to measure as much as possible, but it is a definite limitation of this approach.