Infectious Disease Modelling Flashcards
What is the SIR Model for Infectious Diseases
Susceptible –> Infected –> Recovered
Everyone is allocated to 1 box, can’t be in more than 1.
What is R0
the average number of secondary infections caused by a single typical infectious individual over their lifetime when introduced to a completely susceptible population (depends on both pathogen and population)
What is Re or Rt
(Net/Effective Reproduction Number)
Takes into account the current state of the population:
- not everyone is susceptible (some people are immune etc.)
- changes in contact rate
- If 1 infection is in reality causing <1 case: good.
What is herd immunity
A population is protected when not everyone is immune
(most people are immune so Re (or Rt) is <1).
What is a stochastic model?
Used when there are small numbers of cases - early phase. The outcome isn’t the same each time - can’t precisely predict outcome as get a different answer each time - need many runs to build up a picture of outcomes. Includes variability from chance.
If the population is small, stochastic effects may be of substantial importance and you may never get to a point where stochastic effects are negligible
What is the deterministic model
if you start in the same place with the same parameters you get the same result each time (only applicable to big numbers).
Only 1 run needed for meaningful outcome.
No accounting for probability
Compartmental Models
Population is divided into boxes and the count from each box is modelled.
Deterministic or Stochastic
Easier & less intensive (but can still be complex)
Individual Based Models
aka: infectious disease models, microsimulations, agent based models
Every individual in the population is modelled
Necessarily stochastic (because not just small numbers but 1)
What happens in the growth phase
What is generation time
Exponential Growth.
Generation Time = mean time interval between infection and the onwards infections they cause <– if known can give data about R0.
Need to know R0 and Generation Time because otherwise exponential growth can either be a big R0 and small generation time or a small R0 and big generation time
Contact tracing can help estimate
Turnover
when stops being exponential and starts decreasing (reaches a peak)
SI Model
Susceptible –> Infected
never stop being infected e.g. HIV
If R0 = 2: will stop infecting when 1/2 population is infected.
If R0 = 4 or 5: theoretically will only cause infection when 75-80% of population infected (but in HIV this doesn’t actually happen due to heterogeneity –> a core group e.g. those with lots of sexual partners will have high prevalence but this is usually a small fraction of the population
SIS Model
Susceptible –> Infected –> Susceptible
e.g. Gonorrhoea, Rhinovirus
SIRS Model
Susceptible –> Infected –> Recovered –> Susceptible
Stay immune for a while but it wanes e.g. pertusis
SEIR Model
Susceptible –> Exposed –> Infected –> Recovered
Exposed = latent infection. Allows to build a delay in modelling between infection and infectiousness. Useful for contact tracing and quarantine
What happens after infection spike is over (endgame)
Following a decline & reaching endemicity, children or immigration can increase numbers again. –> the higher the birth rate the smaller the interval between peaks