Summary of "CAUSALIDAD, CRITERIOS DE BRADFORD HILL"

Summary of the Video: “CAUSALIDAD, CRITERIOS DE BRADFORD HILL”

Main Ideas and Concepts

Causality in Medicine The video discusses the concept of causality (etiology) in medicine, emphasizing that there is no strict mathematical formula to determine causal relationships. Instead, causality is inferred based on accumulated evidence and criteria.

Historical Background

Koch’s postulates (1882) were early principles to establish causality for infectious diseases. They required a microorganism to be found in every case of a disease, absent in others, and to cause disease when isolated and introduced to a healthy host.
However, Koch’s postulates represent a simplified view, as many diseases are multifactorial, involving interactions between host, environment, and pathogen.

Association vs. Causation

An association between an exposure and an outcome does not necessarily imply causality.
Confounding variables can create spurious associations (e.g., coffee intake and myocardial infarction risk might be confounded by personality traits).
The video stresses the importance of identifying and accounting for confounders in causal inference.

Examples of Spurious Associations A humorous example is given about chocolate consumption correlating with the number of Nobel prizes in a country, which is likely confounded by socioeconomic factors rather than a causal relationship.

Bradford Hill Criteria for Causality The video focuses on the nine Bradford Hill criteria, which help evaluate whether an observed association is likely causal. These are guidelines, not rigid rules.

Bradford Hill Criteria Explained

Strength of Association
- Measured by relative risk (RR) or odds ratio.
- A stronger association (higher RR) increases the likelihood of causality.
- Example: Helicobacter pylori infection increases peptic ulcer risk significantly (RR ~3 or higher).
Consistency
- Replication of findings across different studies, populations, and times supports causality.
- Lack of consistency does not rule out causality but may reflect study limitations.
Specificity
- A specific exposure leads to a specific disease.
- This is the weakest criterion because many exposures cause multiple outcomes (e.g., H. pylori is linked to ulcers and gastric cancer, and also found in asymptomatic people).
Temporality
- The cause must precede the effect.
- This is essential but can be difficult to establish in cross-sectional studies.
- Example: H. pylori infection precedes ulcer development.
Biological Gradient (Dose-Response)
- Increasing exposure leads to increasing risk.
- Example: Higher antibody levels against H. pylori correlate with higher ulcer risk.
Biological Plausibility
- There should be a reasonable biological mechanism explaining the association.
- Example: H. pylori causes mucosal damage via inflammation, explaining ulcer formation.
- Lack of plausibility may reflect gaps in current knowledge, not absence of causality.
Coherence
- The association should be consistent with existing knowledge and theory.
- Example: Trends in H. pylori prevalence and ulcer/cancer incidence align historically and geographically.
Experimental Evidence
- Randomized controlled trials or interventions that modify exposure and outcome strengthen causal inference.
- Example: Eradication of H. pylori reduces ulcer incidence.
- Not always feasible or ethical (e.g., parachute trials).
Analogy
- If a similar factor causes a similar effect, this supports causality.
- Example: Viral infections or drugs during pregnancy causing birth defects by analogy.

Additional Points

Multicausality Diseases often have multiple causes interacting simultaneously.
Critical Thinking in Research Researchers should question whether their exposure truly explains the outcome or if confounders are responsible.
Avoiding Spurious Associations Statistical significance alone does not imply causality; pathophysiological mechanisms and other criteria must be considered.
Flexibility of Knowledge Scientific knowledge evolves; causal conclusions may change with new evidence.
Practical Application The video encourages applying Bradford Hill criteria to research projects to discern the most important risk factors and avoid misleading conclusions.

Methodology / Steps to Evaluate Causality Using Bradford Hill Criteria

For any observed association, assess the following systematically:

Strength: Calculate relative risks or odds ratios.
Consistency: Look for replication across studies and populations.
Specificity: Determine if the exposure is linked to one disease or multiple.
Temporality: Confirm that exposure precedes outcome.
Biological Gradient: Check for dose-response relationships.
Biological Plausibility: Investigate known or plausible mechanisms.
Coherence: Ensure findings fit with existing scientific knowledge.
Experimental Evidence: Seek intervention or trial data if available.
Analogy: Compare with similar known causal relationships.

Use these criteria as flexible guidelines rather than strict rules. Always consider confounding variables and alternative explanations.

Speakers / Sources Featured

Main Speaker: An unnamed instructor or lecturer conducting a preparatory course in medicine or epidemiology.
Referenced Researchers:
- Bradford Hill (developer of the causality criteria)
- Historical reference to Robert Koch (Koch’s postulates)
- Studies cited from the New England Journal of Medicine and other unspecified epidemiological research.