Validation of three geolocation strategies for health-facility attendees for research and public health surveillance in a rural setting in western Kenya

STRESMAN, G. H.; STEVENSON, J. C.; OWAGA, C.; MARUBE, E.; ANYANGO, C.; DRAKELEY, C.; BOUSEMA, T.; COX, J.

The second method assessed was participatory mapping, and was similar to the recently published 'map-book' exercise [25] and involved producing poster-sized, high-resolution satellite images (Quickbird; Digital Globe, USA) of each facility catchment area (Fig. 2) . Locations of health facilities, schools, markets and other key landmarks were labelled on the image and a reference grid consisting of 2 × 2·5 km 'blocks' was superimposed on the area [27] . Each block comprised 20 'cells', each measuring 500 × 500 m. Each block/cell combination was given a unique numeric identifier. The system (including size of polygon) was selected because it was familiar to the field workers and would provide them a better frame of reference for facilitating the exercise. As part of the participant questionnaire, the interviewer would explain the main features of the satellite map and with the participant, would attempt to locate the residence on the map and record the corresponding cell identifier. Due to the spatial resolution required to locate compounds, this approach is most applicable to local scale but could be scaled up if satellite imagery was indexed into a book-format instead of a poster.