[Spatial analysis of road traffic accidents with fatalities in Spain, 2008-2011].

OBJECTIVE:To estimate the areas of greatest density of road traffic accidents with fatalities at 24 hours per km(2)/year in Spain from 2008 to 2011, using a geographic information system. METHODS:Accidents were geocodified using the road and kilometer points where they occurred. The average nearest neighbor was calculated to detect possible clusters and to obtain the bandwidth for kernel density estimation. RESULTS:A total of 4775 accidents were analyzed, of which 73.3% occurred on conventional roads. The estimated average distance between accidents was 1,242 meters, and the average expected distance was 10,738 meters. The nearest neighbor index was 0.11, indicating that there were aggregations of accidents in space. A map showing the kernel density was obtained with a resolution of 1 km(2), which identified the areas of highest density. CONCLUSIONS:This methodology allowed a better approximation to locating accident risks by taking into account kilometer points. The map shows areas where there was a greater density of accidents. This could be an advantage in decision-making by the relevant authorities.