CONTRIBUTIONS FROM CARTOGRAPHY TO STUDIES ON HEALTH GEOGRAPHY: INVESTIGATING ASSOCIATIONS BETWEEN SPATIAL PATTERNS

Abstract

Identifying patterns in events is the starting point to understand phenomena to which the space and time dimensions are associated in a explanatory way. To illustrate the potentiality of spatial analysis, I evaluate the spatial pattern of mortality from tuberculosis among residents of the Municipality of São Paulo, Brazil, from 2006 to 2009, aggregated by weighting census survey areas. I tested the association between relative risks and spatial clusters and each social indicator as mean monthly income by householder and mean dwellers by residence. Annual mortality rate during studied period was 2.6/100,000 inhabitants. Fifteen spatial clusters were statistically significant, 7 of high risk and 8 of low. The most likely cluster occurred in the downtown area. Although the variable mean monthly income by householder explains only 10% of the spatial distribution of mortality from tuberculosis, almost all high risk spatial clusters overlap weighting census survey areas with mean monthly income below the average of the municipality. Some expected associations traditionally considered evident by epidemiologists explain only a minor parcel of this complexity, leading us to the challenge of integrate different scales and disciplines.