Sampling procedures

For the reason that the study participants are nomadic and scattered, multistage cluster sampling technique was used to select study participants. The clusters were villages with defined geographical boundaries. From a total of 11 kebeles, 6 were randomly selected by a lottery method. The selected 6 kebeles were clustered into 39 villages and from these villages, 17 were selected by systematic random sampling technique. Finally, all (704) households who had U5C were included in the study. For households having more than one child, the younger one was taken as index child and recruited in the study.