Zinc Treatment for Childhood Diarrhea in Bangladesh - Recent Advances and Issues in Environmental Science - page 210

slum or nonslum populations and two districts from each category were randomly
selected. Rural sub-districts were purposively selected on the basis of their proximity
to ICDDR,B field sites in the west (Abhonagar), southeast (Mirsarai), and northeast
(Hobiganj) of Bangladesh. The municipal district capitals representing each rural site
were also selected (Khuna, Camila, and Sylhet). The under-five population of these
ten sites is approximately 1.5 million children.
Figure 1. Summary of the cluster sampling framework for the surveys.
Study Population
Within the selected sites, all wards (defined as urban census tracts or an administrative
grouping of two to three villages) were listed and 20 randomly selected prior to the
start of each survey round. Standard WHO EPI (Expanded Program for Immunization)
cluster methods were applied [9]. These methods eliminate the need to enumerate
and then randomly select households. In rural settings a site central to the village was
chosen and a wheel board spun to determine the direction of households to be visited
following a more or less straight line. Upon reaching the peripheral boundary of the
village the interviewers returned to the central point and spun the wheel again. Within
urban wards a central starting point was chosen and then a systematic, door-to-door
survey of households was carried out. Within each rural or urban ward the survey
was stopped once 16 children, 6-59 months of age with an active or recent (within
the past 2 weeks) episode of diarrhea of at least 48-hour duration had been identified,
and informed verbal consent from either parent obtained. If more than one child in the
household was eligible, one was randomly chosen.