Malaria model predicts successful malaria eradication in regions of mesoendemic transmission

Bob Snow and his colleagues, Ricardo Augas, Lisa White and M. Gabriela M. Gomes published a very interesting study in PLOS One predicting that malaria could be successfully eradicated in regions of mesoendemic transmission (areas with infection prevalence between 11% and 50%, see here) if prevalence of the disease could be brought below a certain threshold. The authors used a classical SIRI-type epidemiological model, where recovered hosts can become infected but not infectious for the second time, and recovered hosts can lose their immunity over time. The authors parameterized this model using clinical data from eight endemic regions in Sub-Saharan Africa. The model predicts a regime of bistability of endemic and malaria-free states, induced by a shorter estimated infectious period for clinical infections, in regions of mesoendemic transmission. These two states are separated by a threshold of prevalence (total number of cases), predicting that malaria can be eradicated in these areas if the number of cases can be kept below a certain level by a combination of interventions. This result is definitely encouraging as both e.g. India, South-America as well as some parts of sub-Saharan Africa belong to the hypo- and mesoendemic region (see this figure). I haven't read the paper thoroughly yet, but it's definitely in my pile. However, just by glimpsing at it, I see that the epidemiological model only follows hosts and not vectors (vectors are implicitly assumed as part of the force of infection). My opinion is that the inclusion of vectors might make the disease dynamics more complex, as well as more realistic. Since the predictions of this paper are so important for public health programmes that aim to control malaria, I would suggest to look at the effect of vectors on the dynamics as well.