Social epi + reflection lecture week 4 Flashcards
What is social epidemiology about?
Social epidemiology = how do social structures, institutions and relationships influence health?
Social epidemiology is often contrasted with risk factor epidemiology. Mention the differences
Social:
social determinants
Upstream, distal conditions
Explains disease risk as due to increased exposure to toxic environments
Leads to macro-level interventions meant to affect population health
Risk factor epidemiology
Clinical, biomed, behavioural determinants
Intra-individual and proximal causal factors
Explains disease risk, poor health at individual level
Leads to individual-specific interventions
What are social determinants of health? What are structural drivers determining these conditions?
Conditions of daily living such as circumstances in which people are born, grown, live, work, age
Education, income , gender, ethnicity, climate, living wage
What are health inequities?
Systematic differences in health status
Arising from social conditions which are unfair and modificable
Difference relative vs absolute inequities?
Relative = social gradient, such as socio-economic position
Absolute = threshold definition (e.g. having degree or not)
What are health determinants?
Biology
Basic needs
Exposure to hazards
Behavior
Access to healthcare
Psychosocial
What is a major challenge in social epidemiology? What is needed?
Showing causation is a major challenge: standard epidemiological methods do not work
We need systems approaches and mixed methods
Why do we use the odds ratio in the cross-sectional study?
there is no baseline incidence or baseline prevalence to compare your selected cases to. Therefore odds ratio. = good for etiologic research, for public health you can use it but not directly calculate the risk difference.
What is the definition + formula of NNT?
1/ARR.
= the number of patients you need to treat to prevent one additional bad outcome
Synonym for random error and what does the random error determine?
Random error = precision. Determines power of a study.
Synonym for systematic error? What can it lead to?
Systematic error = bias. Can lead to false conclusions, can only be estimated
What is considered worse, selection bias or random error in case-control and cohort and why?
Selection bias in cohort study: worse than random error
- Not very likely during recruitment
- Likely to occur as a consequence of loss to follow up. Consequence: over-or underestimation
Selection bias in case-control study: worse than random error
- Selection in cases/controls may depend on exposure
- E.g. non response among controls. Consequence: over-or underestimation
Differential bias: what does it mean?
When exposure and outcome assessment are dependent of each other
Example of differential bias: exposure depends on risk and assessment depends on exposure?
exposure depends on risk (e.g. you had disease, thought about it, think about past diet well. Other person thinks about is superficially and less thought of. Exposure assessment then depends on the risk)
Outcome assessment depends on exposure (being overweight: more inclined to measure blood pressure and blood sugar levels at GP
Differential bias: what happens in your association?
Over- or underestimation
