lecture on hill's criteria

Exam Dates and Structure
  • Exam Availability:
    • Open on Tuesday at the beginning of class and also on Friday.
    • Students can choose to take the exam on Tuesday by informing the instructor as they enter class.
    • The instructor will accommodate by adjusting the date for those who opt to take it Tuesday.
Exam Focus
  • The final exam will cover both theory and math, with a heavier emphasis on theory, as it has not been extensively covered yet.
  • Quantitative Quizzes: These have been introduced in response to prior students' struggles with math-based questions, aiming to balance theory with mathematical understanding.
Hill's Criteria Overview
  • Hill's criteria are utilized to evaluate whether a causal relationship exists between two factors (e.g., exposure and outcome).
    • Descriptive Aspect: Examines who, when, where, and how many are affected.
    • Analytic Aspect: Investigates causal hypotheses regarding why people are getting sick.
    • Key Point: Just because there is an association does not equate to causation.
Key Components of Hill's Criteria

  1. Strength of the Association:

    • Assessed by odds ratio and relative risk.
    • Stronger associations increase the likelihood of causation (e.g., a relative risk of 20 is much stronger than 2).
    • Example: Chimney sweeps in London had a significantly higher risk (82 times more likely) for cancer.

  2. Consistency:

    • The association must be repeatedly observed in different contexts (e.g., populations, times).
    • Consistent results among various studies bolster the case for causal relationships.

  3. Specificity:

    • This implies that a particular exposure causes a specific outcome (though many exposures can lead to various outcomes).
    • For example, smoking is not the sole cause of lung cancer but a significant factor.

  4. Temporality:

    • The cause must precede the effect.
    • Example: If one starts smoking and then gets lung cancer, smoking is not the cause.

  5. Biological Gradient:

    • This suggests a dose-response relationship where more exposure leads to a greater possibility of outcome (e.g., increased cigarettes leading to higher lung cancer rates).

  6. Plausibility:

    • The cause-effect relationship must be biologically plausible given current scientific understanding.

  7. Coherence:

    • The association should fit within existing biological or public health data about the disease.
    • Observation: Males smoke more than females; hence, lung cancer rates should be accordingly higher among males.

  8. Experiment:

    • Randomized control trials or public health interventions provide strong evidence, but ethical considerations often limit feasibility.

  9. Analogy:

    • Comparisons between similar known associations to evaluate new potential associations (e.g., linking thalidomide's effects to known teratogens).
Application of Hill's Criteria: Case Studies
  • Smoking and Lung Cancer:
    • Multiple criteria were evaluated before declaring smoking a causal factor.
    • Example evidence: Mortality rates for smokers are significantly higher (20 to 30 times).
  • Zika Virus and Microcephaly:
    • Strong association was found and evaluated using Hill's criteria, confirming that Zika can lead to significant brain defects.
    • Important findings included that the virus was consistent with outcomes observed from other infectious agents even without a biological gradient due to the binary nature of infection.
Conclusion
  • Evaluating causal relationships requires balancing various criteria. More criteria met can increase the strength of the causal claim, but no magic number specifies how many must be met for confirmation.
  • When studying emerging conditions like Zika, using established criteria helps guide and inform public health policies and recommendations.