Hills Criteria of Causation Flashcards
Strength of Association
Definition: A strong association between the cause and effect is more likely to be causal than a weak one.
Example: A significant increase in lung cancer risk among heavy smokers compared to non-smokers is a strong association.
Consistency
Definition: The association should be consistently observed in different studies and populations.
Example: Multiple studies from various regions consistently show a higher risk of skin cancer among individuals with prolonged sun exposure.
Specificity
Definition: A specific cause should lead to a specific effect. However, this criterion is less emphasized in modern epidemiology.
Example: Certain toxins lead to specific diseases, like lead exposure causing lead poisoning.
Temporality
Definition: The cause must precede the effect in time; the effect should not occur before the cause.
Example: In infectious disease, exposure to a pathogen must occur before the development of symptoms.
Biological Gradient (Dose-Response)
Definition: A dose-response relationship, where an increase in exposure leads to an increase in the risk of the effect, strengthens the evidence of causation.
Example: The more cigarettes a person smokes, the higher their risk of lung cancer.
Plausibility
Definition: There should be a plausible biological mechanism linking the cause and effect.
Example: The link between HIV infection and AIDS is well-understood in terms of the virus’s impact on the immune system.
Coherence
Definition: The association should be compatible with existing knowledge and not contradict well-established principles.
Example: An epidemiological study showing that a commonly consumed fruit is a potent carcinogen would raise questions of coherence.
Experiment
Definition: Experimental evidence, such as randomized controlled trials, can support causation.
Example: Randomized trials show that a new drug effectively reduces blood pressure, demonstrating causation.
Analogy
Definition: Similarities with known causes can provide supporting evidence for causation.
Example: If a new substance has effects similar to a well-known carcinogen, it suggests a potential causal link.
