Skip to content

Cholera WASH Intervention

Jonathan Bloedow edited this page Nov 5, 2024 · 10 revisions

The WASH intervention suppresses environmental transmission. A value of 0 is a no-op and a value of 1.0 is 100% suppression.

We create a scenario with 10 identical nodes with 5 million individuals each, and a CBR of 40. And we turn off contact transmission. We unleash an outbreak at t=40. We also add a "New Infections" reporting channel and plot literally and cumulatively.

With 10 identical nodes we "sweep" by using different values of WASH in each node. First, from 0 to 1.0:

Screenshot from 2024-11-01 17-00-03

Observations: 100% WASH eliminates transmission (cyan flat line) as we'd expect. The other values achieve a range of suppression of incidence as we might expect, though the results are perhaps more logarithmic than linear to the naked eye.

Since most of the discernible effect happens after ~90%, let's "sweep" from 90% to 100%:

Screenshot from 2024-11-01 16-53-22

Below is a non-cumulative plot, first sweeping from 0 to 1.0:

Screenshot from 2024-11-01 17-02-12

Then from 0.9 to 1.0: Screenshot from 2024-11-01 17-14-33

and then zoomed in on the last two recurrent outbreaks which shows differentiation between WASH settings (0.9 to 1.0): Screenshot from 2024-11-01 17-14-53

Mid-Sim Intervention

Finally, while it's expected that a geographical region -- represented as a node or patch -- will have the same WASH value for the whole sim, we may want to test interventions which modify (usually increase) the WASH parameter. In the test below, we used a full SSA-like setup -- 41 nodes with varying populations and birth rates, and seasonality from data -- and we start with WASH values all driven from data, and then half way through we set every other node/patch to 100% (1.00). I disabled contact transmission just to make the effect more obvious. (The height of the red/prev curves seem shorted because the y axes adjusted for the fact that the S values go higher than before after the perfect WASH goes into effect). The purpose of doing every other node is to demonstrate that the mid-sim WASH change can be set spatially, not just globally.

image

Next we set WASH to 0 (off) for the first half of the sim, and then switch to 1 (on) everywhere for the second half. First with full transmission...

image

.. and then environmental transmission only: image

Clone this wiki locally