BIOSTATS LEC CAUSALITY

• Epidemiology


  study of the distribution of determinants of health-related states or events in specified populations, and the application of this study to control of health problems
• Causality


  an event, condition, or characteristic that preceded the outcome or disease event and without which the event either would have not occurred at all or would have not occurred until some later time
• Direct cause


  A factor that causes the problem without any intermediate steps
• Direct cause


  Ex. Broken leg due to car accident
• Indirect cause


  A factor that may cause the problem but with an intermediate factor or step
• Indirect cause


  Ex. Alcohol causes car accident which causes broken leg

  Alcohol is indirect cause
• Bidirectional cause


  Each of two variables may reciprocally influence the other
• Koch's Postulates


  Causality of Infectious Diseases
• 4 rules that establish the causal relationship between an infectious agent and a particular disease
• 4 rules that establish the causal relationship between an infectious agent and a particular disease

  1. An organism can be isolated from a host suffering from the disease
  2. The organism can be cultured in the laboratory
  3. The organism causes the same disease when introduced into another host
  4. The organism can be re-isolated from that host
• Epidemiologic Triad


  Interplay of different factors: Susceptible Host, Pathogenic Agent, Environment
• Interplay of different factors

  • Susceptible Host
  • Pathogenic Agent
  • Environment
• Multifactorial Model

  • Sufficient cause
  • Necessary cause
• Sufficient Cause


  A set of factors whose completion inevitably leads to the outcome
• Necessary cause


  A factor present in every sufficient cause
• Necessary and Sufficient


  Multifactorial Model

  
• Necessary, But Not Sufficient


  Multifactorial Model

  
• Sufficient, But Not Necessary


  Multifactorial Model

  
• Neither Sufficient, Nor Necessary


  Multifactorial Model

  
• Association

  • Identifiable link between two variables
  • Does NOT readily imply cause-effect relationship
  • Statistically significant
• Evaluating Causation

  1. Investigation of Statistical Association
  2. Investigation of Temporal Relationship
  3. Elimination of All Known Alternative Explanations
• Investigation of Statistical Association

  • Risk factor
  • Protective factor
  • Confounder
• Risk factor

  • An exposure, behavior, or attribute that, if present and active, clearly increases the probability of a particular disease occurring in a group of people compared with an otherwise similar group of people who lack the risk factor
• Risk factor


  Neither a necessary nor a sufficient cause of disease
• Protective factor


  Vaccine
• Protective factor


  Less likely to be found among those with the disease
• Confounder


  Extraneous variable whose effect influences the relationship of the exposure and outcome of interest
• Confounder


  Third carriable problem
• Confounder


  Has a direct effect
• Investigation of Temporal Relationship

  • Experimental Study
• Experimental Study


  Randomization
• Experimental Study


  Measure the risk factor and disease in both groups
• Experimental Study


  Before and after the experimental intervention
• Elimination of All Known Alternative Explanations


  No other likely explanations for the association
• Bradford Hill


  Guideline for judging whether an observed association is causal
• Guideline for judging whether an observed association is causal

  • Long latency pd and possibility that there may be multiple causes for the same disease
  • Useful to remember distinctions between association and causation
  • Designing epidemiologic studies
  • Interpreting the result
• Temporal relationship


  Exposure (agent/risk factor) always precedes the outcome
• Strength of association


  The size of the association as measure by appropriate statistical tests
• Dose-response relationship


  An increasing amount of exposure increases the risk
• Consistency


  The association is consistent when results are repeated in studies in different settings using different methods