Sir Austin Bradford Hill (1897–1991) was a British epidemiologist and statistician. He helped establish the modern randomized controlled trial.
In 1965, he delivered a landmark address titled The Environment and Disease: Association or Causation? Observational studies often show associations, but association alone does not prove that one factor causes another. Hill asked: when might an association support a causal interpretation?
He proposed nine viewpoints to guide that judgment, with many adding reversibility as a tenth consideration.
Tap each cell under Early Antibiotics and Sepsis to reveal that criterion’s application to the vignette.
| # | Criterion | One-line meaning | Early Antibiotics and Sepsis |
|---|---|---|---|
| 1 | Strength | Large association | Strong — reported hourly mortality increases are often large |
| 2 | Consistency | Seen across studies and settings | Moderate — most cohorts show benefit, though effect sizes vary |
| 3 | Specificity | Exposure linked to a specific outcome | Weak — sepsis etiology is varied and mortality has many causes |
| 4 | Temporality | Exposure precedes outcome | Moderate — antibiotics precede death, but time zero is often fuzzy |
| 5 | Biological gradient | Dose–response relationship | Strong — more delay associated with worse outcomes (if real) |
| 6 | Plausibility | Fits known biology | Strong — infection treated with antibiotics is biologically plausible |
| 7 | Coherence | Fits broader evidence | Strong — aligns with animal model data, sepsis bundles, and Surviving Sepsis guidance |
| 8 | Experiment | Intervention proves causation | Weak — RCTs of intentional delay are unethical; evidence is mostly observational |
| 9 | Analogy | Similar exposures behave similarly | Moderate — other time-sensitive ED care (STEMI, stroke) follows similar logic |
| 10 | Reversibility (often added) | Removing exposure reduces outcome | Weak — delaying antibiotics worsens outcomes in models, but evidence remains observational |