Mapeo de enfermedades con enfoques bayesianos y evaluación del riesgo relativo para la salud pública en Colombia
Fecha
Autores
Autor corporativo
Título de la revista
ISSN de la revista
Título del volumen
Editor
Compartir
Director
Altmetric
Resumen
In Colombia, the data on the presence of vector-borne diseases have an over-dispersion problem, which means that more than 70% of the municipalities do not have reported cases of the diseases or simply do not enter them in the SIVIGILA application of National Institute of health, also given the low operation in the health system in the country, case records for some municipalities can be concentrated in large cities, this means that cases registered in large municipal areas such as Bogotá do not really they have a count of the disease, if not this count is part of the nearby areas with deficiencies in the health system where this disease actually occurs. Thus, the estimation of the risk that a municipality in Colombia may have of a disease can be determined using the INLA algorithm taking into account these spatial relationships and the presence of outliers given the inefficient health system. Bayesian approaches have the quality that not only is modeling carried out on the data in question, but also that it takes into account the information that an experienced researcher or a previous study can provide to the project. Within the spatial associations, spatial weighting matrices (W) were determined that will minimize the correlation. In addition, models with different probability distributions associated with counts with specific treatments to the over-dispersion of the data are established.