Fix bug with pev min_wait

[giovannic]

This PR addresses an issue where min_wait is apparently ignored. It is easiest to see when you have two mass_pev campaigns close to each other (example below). But would likely affect other combinations of pev interventions (including those using EPI).

library(malariasimulation)

p <- get_parameters(overrides = list(human_population = 10000,
                                     model_seasonality = TRUE,
                                     individual_mosquitoes = FALSE))

p <- set_mass_pev(
  parameters = p,
  profile = rtss_profile,
  timesteps = c(20, 100),
  coverages = c(1, 1),
  min_ages = 0,
  max_ages = 100*365,
  min_wait = 500,
  booster_timestep = 100, # timesteps following initial vaccination
  booster_profile = list(rtss_booster_profile),
  booster_coverage = 1
)

o <- run_simulation(timesteps = 365,
                    parameters = p)

print(sum(o$n_pev_mass_dose_1))
print(o$n_pev_mass_dose_1)

The problem was that we were sampling pev populations based on the 3rd dose instead of the 1st. This PR corrects that while maintaining the previous behaviour wrt vaccine efficacy.

-- Changes --

Fix bug with pev min_wait:

• Update pev-epi min_wait test to directly check sample population
• Update pe-epi min_wait test to check that pev_timestep is being updated on the first dose
• Copy pev-epi min_wait test to mass pev
• Update pev and epi processes to calculate targets based on time of first dose
• Update infection immunity function to calculate pev immunity only after 3rd dose

[kellymccain28]

This is great. Thanks for fixing this Giovanni!

I tested with multiple mass vaccination campaigns and with a combination of EPI and mass campaigns - all seemed to work as expected.

[pwinskill]

Thanks @giovannic!

I'm being slow and stuggling to follow the logic through all the way. Can I check that this change still has efficacy starting at the third dose (whist changin min_wait to respect time since first dose)?

I'm wondering if it would be clearer to implement with an additional variable(s) instead of the resetting of variables. Something like:

variables$doses = ...
variables$time_of_first_dose = ...
variables$time_of_third_dose = ...

Is that overkill?

[giovannic]

Thanks @giovannic!

I'm being slow and stuggling to follow the logic through all the way. Can I check that this change still has efficacy starting at the third dose (whist changin min_wait to respect time since first dose)?

I'm wondering if it would be clearer to implement with an additional variable(s) instead of the resetting of variables. Something like:

variables$doses = ... variables$time_of_first_dose = ... variables$time_of_third_dose = ...

Is that overkill?

Not resetting pev_profile is a good point. Not that anyone is modelling this now, but this change would lead to weird behaviour if there were two mass_pev campaigns in a simulation. Efficacy for the first campaign would be reset on enrollment to the second campaign.

---

I'm not against your alternative.

Would the events change too? If the events remain the same, there'd be a lot of redundancy between the variables and the events. Currently, pev events fire on every dose so variables$doses and variables$time_of_third_dose would not be needed.

Also, how would we deal with booster timesteps?

---

Perhaps a clearer way would be to have:

variables$last_pev_timestep, variables$last_efficacious_pev_timestep and variables$pev_efficacy_profile?

Then we could set variables$last_pev_timestep on the first dose; and then all three of the variables on the third dose and boosters.

[pwinskill]

Thanks Giovanni - that sounds sensible to me. I can certaintly see the sitation with multiple mass pev campaigns arising so think we should be robust to that.

[giovannic]

@pwinskill I've updated this PR to have an explicit tracking of (non-)efficacious doses. I've updated the existing tests to check the new behaviour and that seems to be enough, but let me know if we should add some more!

[pwinskill]

Thanks very much @giovannic. I can follow the logic much more clearly now - much appreciated 🙏