Alls,
I'm working on the distribution of Yellow fever human cases (ongoing outbreak) in Brazil and expect to use, among other, Ecological Niche Modelling.
I'm using a set of "classical" biotic and abiotic variables (some bioclim, human footprint, Biodiversity Intactness Index, mammal richness, srtm, canopy, vegetal biomass, human density). The map of predictions of more "favorable" region is expected to be a map of risk (under epidemiological definition).
I have a concern with the vaccination coverage. Would you use this as an other putative explanatory variable, during the model building, together with the above mentionned ? Or rather, should we keep this information for discussion of the risk map, more to weight this risk once "ecologically" (sensu niche mdelling definition) identified ?
Comments welcome !
There could be several ways to go about this, including the two you've mentioned. However, if you're looking at vaccination coverage, you'd probably want the data to be spatially resolved at scales comparable to your other variables for it to offer an explanation of the variability in risk within Brazil. If this is the case, then you could have a nice way to simulate threshold levels of coverage to modify risks.
Or, you could look at it more coarsely/categorically with stronger assumptions, i.e., urban regions have at least 60% coverage vs. rural regions which have below 60%. In the latter case, human density may serve as a weak proxy for vaccination coverage, and perhaps it's an effect that you won't be able to tease apart. Then I'd consider using this information as a discussion point.
In any case, timely and important work.
Good luck!
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology