P. Escobar-Hernández, A. López-Quílez, M. Marco, M. Montagud-Andrés, E. Gracia, M. Lila
Multivariate spatial disease mapping has been on the rise in importance over the last years. The recent advantages in computational capacity and the increasing amount of data available has allowed epidemiologists to start analysing the share effect of several outcomes. In this context, and from a Bayesian perspective, the emergence of the Integrated Nested Laplace Approximation (INLA) has been a complete revolution, allowing statisticians to perform complex models that were unaffordable one decade ago. This presentation focuses on comparing the performance of several types of multivariate spatial models that can be implemented within the R-INLA and INLAMSM packages. Initially, the comparison is based on a simulated dataset. Furthermore, we analyse a dataset consisting on suicide-related emergency calls, grouped by gender, type of caller (victim vs witness), age groups and time period to study the effect that COVID lockdowns could have had on mental health in Comunitat Valenciana.
Palabras clave: Spatial, Multivariate, Disease Mapping, Epidemiology
Programado
Bioestadística
8 de noviembre de 2023 17:20
HC4: Sala Sacristía