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.

To facilitate participatory mapping, a grid was superimposed onto the study area, leading to an edge effect whereby if a person was located just outside of the block/cell they would be classified incorrectly even though the error margin could be only a few metres. Edge effect will always be an important limitation that must be accounted for in any application of this methodology particularly when the focus is on locating residences at a precise spatial resolution. However, despite this limitation, this research has provided important insight into how the edge effect can be minimized and sensitivity increased by the addition of buffer zones. Other approaches could have been used including a hexagonal grid or larger clusters as was used in the study in Blantyre's urban slum area [25] . These approaches will likely reduce, but not completely eliminate the edge effect. Moreover, in this study, there was a significant difference in the proportion of people correctly located at each health facility and not every participant was willing to complete the exercise. This suggests that the familiarity of the interviewers with the area, their ability to read and explain the maps to local populations, and the time they have or choose to dedicate may be important determinants for success.