Skip to content


Basic Definitions

attack rate: the % of people who get sick after exposing to something

relative risk: the % who become ill in one group divided by the % who become ill in another group, i.e. the ratio of two attack rates

case & control 实验组与对照组?

incidence & prevalence: Thinking about a tank of prevalent cases, then the incident cases shoud come in while dealth or recovery should come out.

prevalence: number with the disease/population size

prevalence depends on incidence rate and duration:

incidence: number of new cases/population size at risk

case-fatality ratio(CFR): the percent of people with the condition who die in a time period (death/cases)

age-adjusted death rate:

life expectation

Sometimes there is no midpoint and we will use the upper bound of each age group.

disability-adjusted life year(DALY): the day you could live times your life quality (from 0 to 1)

When there will be multi-possible outcomes, use:

potential years of life lost(PYLL):


R0 - the number of people that one person could spread to

We could reduce R0 by:

  • vaccinating and reduce the number of people who are susceptible
  • quaranting, to reduce transmission (or culling animals)

SIR Model

  • susceptible
  • infected
  • Recovered or removed

S + I + R = 1

Given transmission rate and recovery rate :

SEIR Model

  • susceptible
  • exposed
  • infected
  • recovered

Or, you have to add births, deaths, migration or carriers of a disease.

After adding death parameter =birth parameter , where is the death possibility of the disease:

To stop an epidemic, we need R0<1.

  • make S small by vaccinating
  • make small by washing hands or social distancing
  • make large by quarantine(shorten)


herd immunity: we only need to vaccinate a portion of people to eradicate an infection, the portion is:


incubation period: the time between infection and onset of clinical disease/symptoms

latent period: time between infection and becoming infectious