Revista Española de Salud Pública
Print version ISSN 1135-5727
GOMEZ-BARROSO, Diana et al. Space distribution of tuberculosis in Spain by geostatistical methods. Rev. Esp. Salud Publica [online]. 2009, vol.83, n.5, pp. 737-744. ISSN 1135-5727. http://dx.doi.org/10.1590/S1135-57272009000500014.
Background: Tuberculosis incidence has been associated with many factors, both epidemiological and social. In Spain, tuberculosis is a statutorily notifiable disease requiring individualised reporting. During the last few years rates of respiratory tuberculosis show a steady decline. This study sought to assess respiratory tuberculosis morbidity and mortality in association to socio-economic and epidemiological covariates and estimate its spatial distribution across the country, using geo-statistical methods. Methods: Respiratory tuberculosis incidence rates were standardised by age and sex with the data of the National Epidemiological Surveillance Network (RENAVE, Red Nacional de Vigilancia Epidemiológica) for 2006. The following socio-economic variables were included in the study: socio-economic status, educational level, overcrowding rate, population density, standardised immigration rate by sex, unemployment rate and average spending per person in euros. The epidemiological variables included were, such as, AIDS rate and the influenza incidence rate. To assess the association of covariables a multivariate analysis was performed using a Generalised Linear Model assuming Poisson distribution. The goestatistical method co-kriging was adjusted with the significant variables to built the spatial distribution of risk. Results: The statistically significant covariates were overcrowding rate, standardised immigration rate by sex, educational level, unemployment rate, average spending per person in euros, AIDS rate and the influenza incidence rate. The geostatistical method shows spatial variability of the risk with higher risks in the northwest and southeast of the peninsula. Conclusion: Results prove that the co-kriging method is a useful tool to show the spatial distribution of risk. Alternatively, tuberculosis is associated with both social and epidemiological covariates.
Keywords : Tuberculosis; Spatial analysis; Multilevel Analysis.