ModelIntervMisc

Pseudo interventions

The simulation description can include a number of actions that, while not technically corresponding to malaria interventions, act on the simulated population in an analogous way and are hence specified in the intervention section of the scenaria description. The following table lists these pseudo-interventions and the deployment methods for them:

pseudo-intervention	mass-deployment	continuous-deployment
Import infections	yes	no
InsertR_0Case	yes	no
immuneSuppression	yes	no
recruitmentOnly (sub-populations)	yes	yes

Import Infections

This pseudo-intervention provides a way to model infections being imported into the simulated population. It is based on an importation rate, x(t), and forcibly importing infections according to the Bernoulli distribution (that is, assuming an importation rate of r infections per person per time-step, for each simulated human and for each time-step the probability of that person receiving an imported infection is r).

The importation rate is actually specified with units of imported infections per 1000 people per year. In simulations with a change in importation rate, the initial importation rate (at time=0) should be specified, even if this is 0. In the following example, 5 infections per 1000 people per year are imported for the first 10 years, after which the importation rate doubles:

<interventions>
  <importedInfections name="example">
    <timed>
      <rate time="0" value="5"/>
      <rate time="3650" value="10"/>
    </timed>
  </importedInfections>
</interventions>

If you only need a constant infection-importation rate, only include one <rate> element with time 0. Periodically varying importation rates can also be specified, for example using three different rates depending on the time of year (this example using the one day timestep):

<importedInfections>
  <timed name="annually varying rates" period="365">
    <rate time="0" value="1"/>
    <rate time="100" value="5"/>
    <rate time="200" value="2"/>
  </timed>
</importedInfections>

Insert R_0 case

As of schema 32, this intervention has been removed. If necessary it can be re-introduced.

The InsertR_0Case pseudo-intervention can be used in estimating reproduction numbers. It is assumed before deployment that all infections have been removed from the population, and that all humans are completely protected with a transmission-blocking vaccine. To estimate the Basic Reproduction Number, R₀ theimmuneSuppression intervention is also applied to remove all natural immunity.

The InsertR_0Case pseudo-intervention randomly selects one human, removes his/her transmission-blocking vaccine but provides the equivalent of a pre-erythrocytic vaccine to prevent re-infection (bypassing the usual vaccine configuration to allow this person to have a PEV vaccine without the rest of the population), and finally infects this human. Since only one person is infected and is able to transmit infections, the total number of infections received after this (reported by "new infections" output) is the number of secondary cases caused by one primary infection.

The XML description accepts no configuration other than time:

<insertR_0Case>
  <timed time="146"/>
</insertR_0Case>

Immune suppression

The immuneSuppression intervention removes all natural immunity (though not vaccine effects) from humans. It does not prevent development of new immunity. This functionality can be used in estimating the Basic Reproduction Number, R₀.

It is described with standard timed deployment:

<immuneSuppression>
  <timed coverage="1" time="78"/>
</immuneSuppression>

Recruitment only

To allow recruitment into a sub-population without an associated intervention deployment, the pseudo-intervention recruitmentOnly can be used:

<interventions>
 </human>
  <component id="my sub-pop">
    <recruitmentOnly/>
  </component>
  ...
  <deployment>
    <component id="my sub-pop"/>
    <!-- continuous and/or timed deployments here to recruit members -->
  </deployment>
 </human>
</interventions>

Note: if deploying an intervention to a sub-population on the same timestep as that sub-population was recruited, read order of deployments.

Change EIR

The changeEIR intervention is used to override the default transmission settings. It can only be used with the non-vector transmission model. It is used to simulate the impact of an intervention package with known impact on EIR, either on public health outcomes, or on as a factor modifying the impacts of interventions that can be simulated with the non-vector transmission model. It can be used, for example, to switch to EIR settings recorded during a trial, while retaining the main transmission description for the warm-up and pre-trial periods of the simulation. This intervention updates the transmission model with an entirely new description. It is described like:

<interventions>
  <changeEIR name="name of new transmission settings">
    <timedDeployment eipDuration="10" time="0">
        <EIRDaily origin="interpolate">0.00219</EIRDaily>
        <EIRDaily origin="interpolate">0.00210</EIRDaily>
        <EIRDaily origin="interpolate">0.00202</EIRDaily>
        ...
      </timedDeployment>
  </changeEIR>
  ...
</interventions>

The new EIR is always matched to the whole intervention period, thus must contain enough entries to cover from the beginning of the intervention period to the end, even if not deployed at time 0 or replaced before the end of the simulation. Values are, however, only used from the time of deployment (with some values potentially going unused).

Any number of timedDeployment elements may be used, each of which specifies a time of deployment (in timesteps) and has the same type of content as the nonVector element used to describe the initial transmission.

Deploying a <changeEIR> intervention changes the transmission mode parameter to force EIR (human infectiousness to mosquitoes has no further effect on EIR).