Table 1.
Features extracted for all the analyses by state used to predict the mortality risk from COVID-19 among tested individuals in Mexico.
Table 2.
Fig 1.
Maps associated with COVID-19 deaths in Mexico by state until April 21st, 2020, adjusted for sex and age.
A) Quartiles corresponding to mortality risks among tested individuals smoothed through an empirical Bayes procedure. B) Standardized mortality ratio. Shapefile from http://tapiquen-sig.jimdo.com under a CC BY license, with permission from Carlos Efraín Porto Tapiquén, original copyright 2015. Note: The quantities in brackets beside the categories correspond to the number of states in each category.
Fig 2.
Spatial clustering associated with mortality risk from COVID-19 among tested individuals adjusted for sex and age in Mexico by state until April 21st, 2020, considering queen contiguity.
A) Significant spatial clustering obtained through Local Indicators of Spatial Autocorrelation (LISA) comparisons. There are four types of clusters: High-High, Low-Low, High-Low, and Low-High, e.g. a Low-Low cluster (blue) indicates states with low values of a variable significantly surrounded by regions with similarly low values. B) P-values associated with the spatial clustering in A), C) Scatter plot associated with the smoothed risks vs their corresponding spatially lagged values, including the associated linear regression fitting, whose slope is the Moran’s I statistic. Shapefile from http://tapiquen-sig.jimdo.com under a CC BY license, with permission from Carlos Efraín Porto Tapiquén, original copyright 2015. Note: The quantities in brackets beside the categories correspond to the number of states in each category.
Fig 3.
Figures associated with the selection and goodness-of-fit of a multivariable generalized geographically weighted model (GGWR), with a poison distribution, offset, and a logarithm link function explaining the mortality risk from COVID-19 among tested individuals.
A) Correlation plot including the raw risks and B) Representation of Pearson residuals by state. Shapefile from http://tapiquen-sig.jimdo.com under a CC BY license, with permission from Carlos Efraín Porto Tapiquén, original copyright 2015. Notes: (1) Offset: Log of the total number of people tested in a state; (2) Multiplicative changes in mortality risks (MRt) from the effects of different risk factors by state.
Table 3.
Statistics by variable (minimum, maximum, and quartiles) corresponding to the effects by state* over the mortality risk from COVID-19 among tested individuals (MRt) under the GGWR and similar effects and p-values associated with a global model (all models consider a Poisson distribution, offset term**, and logarithmic link function).
Fig 4.
Multiplicative estimated effects over the tested mortality risk due to COVID-19 among tested individuals (MRt) under a GGWR for those variables that are significantly associated with the response under a global model (Poisson models with offset and a logarithmic link function).
Shapefile from http://tapiquen-sig.jimdo.com under a CC BY license, with permission from Carlos Efraín Porto Tapiquén, original copyright 2015. Note: (1) Offset: Log of the total number of people tested in a state; (2) Multiplicative changes in mortality risks (MRt) from the effects of different risk factors by state.