details on hills criteria and shortnote
Reading File
Finding Sources
Finding Sources
Searching the Web
Bradford Hill criteria causation epidemiology 9 criteria detailed explanation
Reading File
Reading File
Here is a detailed explanation of Hill's Criteria (Bradford Hill's Viewpoints for Causation):
Hill's Criteria of Causation
Proposed by Sir Austin Bradford Hill in his landmark 1965 paper "The Environment and Disease: Association or Causation?" (Proc R Soc Med. 1965;58:295-300), these nine viewpoints help evaluate whether an observed statistical association between an exposure and a disease is likely to be truly causal.
Hill himself called them "nine viewpoints from all of which we should study association before we cry causation" - NOT a rigid checklist. No single criterion is sufficient or necessary (except temporality).
The 9 Criteria (Mnemonic: S-C-S-T-B-P-C-E-A)
| # | Criterion | Explanation |
|---|---|---|
| 1 | Strength of Association | The stronger the association (higher RR/OR), the more likely it is causal. Example: smokers have 9-10x risk of lung cancer. |
| 2 | Consistency | The association is repeatedly observed in different populations, places, times, and by different researchers. |
| 3 | Specificity | One cause produces one specific effect. If a factor is associated with only one particular disease and in specific populations/sites, causation is more likely. |
| 4 | Temporality | The cause MUST precede the effect. This is the only absolute requirement - exposure must come before disease. |
| 5 | Biological Gradient | A dose-response relationship: increasing exposure leads to increasing disease risk; reducing exposure reduces risk. |
| 6 | Plausibility | The association makes biological sense and is consistent with known mechanisms of disease. |
| 7 | Coherence | The association is compatible with existing knowledge, natural history, biology, and epidemiology of the disease. It does not conflict with known facts. |
| 8 | Experimental Evidence | If manipulation of the exposure (e.g., removing the factor via RCT or natural experiment) alters the outcome, causation is supported. |
| 9 | Analogy | A similar cause-effect relationship is already established. If one drug causes birth defects, a similar drug might too. |
Short Note: Bradford Hill's Criteria
Definition: A set of 9 epidemiological viewpoints used to assess whether an observed association between a risk factor (exposure) and a disease (outcome) is causal in nature.
Background:
- Proposed in 1965 by Austin Bradford Hill, a British epidemiologist and statistician.
- Originally applied to the association between cigarette smoking and lung cancer.
- Later adapted for observational studies where RCTs are not feasible.
Key Points to Remember:
- These are guidelines, not rules - no single criterion is obligatory (except temporality).
- A stronger causal argument requires satisfying more criteria.
- Temporality is the only criterion considered essential/absolute - the cause must always precede the effect.
- Strength and consistency are considered the most important in practice.
Common Mnemonic: "Stronger Consistency Seals The Biological Plausibility, Coherent Experiments Allow"
Or simply: SCSTBPCEA
Limitations:
- Does not apply well to multifactorial diseases (e.g., cardiovascular disease has many causes).
- "Specificity" is considered the weakest criterion in modern epidemiology.
- Cannot definitively prove causation - they only support or weaken the causal argument.
- Not a substitute for RCT evidence.
Classic Example - Smoking and Lung Cancer:
- Strength: Smokers have ~9x risk vs. non-smokers.
- Consistency: Confirmed across many countries and studies.
- Temporality: Smoking precedes cancer development.
- Dose-response: More pack-years = higher cancer risk.
- Plausibility: Carcinogens in tobacco are known mutagens.
- Experiment: Cessation reduces cancer risk over time.
10th Criterion (sometimes added):
Some sources add Reversibility - removal of the suspected cause leads to improvement or disappearance of the effect. This is especially relevant in pharmacovigilance (drug adverse effects).
Sources:
- Brenner and Rector's The Kidney (Table 19.2: Bradford Hill Criteria for Causation), adapted from Hill AB. Proc R Soc Med. 1965;58:295-300
- Fitzpatrick's Dermatology, Vol 1-2 - Hill's criteria in the context of observational studies
- Wikipedia - Bradford Hill Criteria
- PMC - Applying Bradford Hill Criteria in the 21st century
This is a shared conversation. Sign in to Orris to start your own chat.