Module 2 Flashcards
Epidemiology
Study of distribution and determinants of health-related states or events in specified populations and application of this study to control of health problems
Importance of pop. Health
Lifestyle, environment and public health improvements play significant roles in health + mortality
Aim of public/population health framework
Provide max. Benefit for largest no. People at same time reducing inequities in distribution of health and wellbeing
“Define problem” study
Cross-sectional
“Identify risk and protective factors” study
Cohort and case-control
“Develop and test prevention strategies” study
RCT and diagnostic test accuracy
“Assure widespread adoption” study
Evaluative studies
Preventative action can be
Put in place before identifying causative factor
- knowledge of complete pathway is not a pre-requisite for introducing preventative measures”
- can help reduce disease occurrence in pop.
James Lind’s experiment
One of earliest controlled trial although patient no. Were very low
- example of prevention before identifying cause (vitamin C deficiency)
Causal relationships
Epidemiology examines relationships/association between exposures and outcomes for this purpose but remember correlation/association doesn’t always = causation
Determining causality
- can NOT be proven in human studies (practical and ethical reasons)
- most non-experimental in ‘noisy’ environments thus must beware of errors
Bradford Hill framework - aid to thought
1) temporality
2) strength of association
3) consistency of association
4) biological gradient/dose-response
5) biological plausibility of association
6) specificity of association
7) reversibility
Temporality
Cause THEN outcome
- easier in cohort than cross-sectional/case-control studies
Strength of association
Measured by size of relative risk
Dose-response
Incremental change in exposure = change in disease rates
- linear dose-response relationship
Specificity
Cause = single effect or single cause = effect
- weakest feature as health issues have multiple interacting causes and many outcomes share causes
Reversibility
Under controlled conditions (RCT), exposure change = outcome change
(Cause deleted = outcome deleted)
- strongest evidence but not always possible
Rothman’s causal pie components
Recognises multicausality
- sufficient cause
- component cause
- necessary cause
Sufficient cause (causal mechanism)
Whole pie
- min. Set of conditions
- often several factors
- 1 disease may have several sufficient causes
Component cause
Slice
- contributes towards disease causation
- insufficient alone
- interact to produce disease
Necessary cause
A component cause that MUST be present for specific disease to occur
- some diseases may not have one so a component cause will be a necessary cause
Prevention using causal pie
Blocking/removing any component cause = prevention of some cases of disease
(No need to identify every component cause)
- can intervene at any number of points in pie
Causal pie limitations
- fails to capture dose-response relations as a continuum (just series of discrete sufficient causes)
- assumes all causes are deterministic (occurrence completely determined by combo of causes without randomness)
Probabilistic concept of causation
- cause increases probability/chance that its effects will occur
- sufficient cause raises prob to 1
- necessary cause raises prob from 0
- each component cause contributes towards prob from 0 to 1
- considers environmental factors, group level effects