17. Epidemiological indicators II: definition and measures of relative risk and odds ratio

From greek.doctor
  • Risk is a proportion; odds is a ratio
  • Risk
    • = the probability that an event will occur
    • Risk = number of times an event occurred / total number of possible occurrences
    • Estimated based on incidence
    • If there are 10 people with cancer, and 6 die, the risk of dying is 6/10 = 60%
  • Odds
    • = the chance that an event will occur
    • Odds = number of times an event occurred / number of times the event didn’t occur
    • If there are 10 people with cancer, and 6 die, the odds of dying is 6/4
  • Indicators of association between the disease and its risk factors
    • Relative risk
    • Odds ratio
    • Proportional mortality ratio
    • Standardized ratio
  • Relative risk (RR)
    • The proportion of risk for a particular disease in the risk factor-exposed and non-risk factor-exposed population
    • Basically how much higher the risk of disease is in those who are exposed to a certain risk factor
      • If RR > 1 there is a positive association between factor and the disease, so it is a risk factor
      • If RR = 1 there is no association between factor and the disease
      • If RR < 1 there is a negative association between factor and the disease, so it is a protective factor
    • RR indicates the strength of the association
    • Example from John Snow’s water pump:
      • People who drank water from the pump had 58,4% risk of cholera, while those who didn’t had only 6,7%
      • The relative risk is 0,584/0,067 = 8,72. The risk of cholera was almost 9 times higher in those who drank from the pump
    • Used in cohort studies
  • Odds ratio (OR)
    • The ratio of the odds of having the disease in the population exposed to the risk factor and the odds of having the disease in the non-exposed population
      • If OR > 1 the disease is more likely in exposed than non-exposed population
      • If OR = 1 the disease is equally likely in exposed and non-exposed population
      • If OR < 1 the disease is less likely in exposed than non-exposed population
    • Used in case-control studies to measure the strength of association between risk factor and disease
      • In case-control studies, relative risk can’t be calculated. However, for rare diseases, the OR is almost equal to the RR and is therefore used to estimate it