Abstract

BACKGROUND:

Clusters of infection can indicate the underlying risk pattern of an endemic disease. Retrospective epidemiological data have been used to map the risk of tick-borne encephalitis (TBE) and Lyme borreliosis (LB) in the Central Bohemian region of the Czech Republic.

METHODS:

Both reported places of infection and patients' residences were entered in a geographical information system; their distance distribution and census data were used to model density of the population at risk. Point-pattern analysis and non-parametric kernel smoothing of points of infection were applied to compute the risk maps. Tick flagging and direct immunofluorescence assay were used to probe true LB-risk in the field.

RESULTS:

Tick-borne encephalitis infections proved to be more clustered than those of LB which was widespread; however, the most prominent clusters of both diseases largely correspond to each other. The estimated LB risk correlated well with tangible disease challenge as assessed from the tick abundance and Borrelia infection rates at 15 selected localities surveyed annually.

CONCLUSION:

The risk of LB is widely and smoothly distributed over the area studied, apparently following tick habitats wherever they occur, while TBE is confined to a subset of these locations.