Purpose of public health
- measure disease burden in a population
- healthcare planning and policy
- identify patterns/occurrence of disease
- identify causes
- evaluating interventions
Advantages and disadvantages of cross sectional studies
Advantages
- quick
- relatively low cost
- multiple causes of disease can be studied
- multiple consequences of a given exposure can be studied
- disease prevalence can be measured
Disadvantages
- subject to recall bias as they have to remember past exposures
- confounding factors may influence results
- cannot use it to estimate relative risk directly
- not suitable for rare disease
- statistically insufficient for studying a single disease - can identify associations but cannot infer causality
- may be misleading if exposures change over time
Advantages and disadvantages of ecological studies
Advantages
- low cost because data is already available
- generates a hypothesis to be tested by other study designs
- gives exposure information at the group level that may not exist at the individual level
Disadvantages
- measure of exposure is based on averages for the population
- data on confounders is often unavailable
- population exposures may not reflect individual exposure levels
What is the ecological fallacy?
Observations about the whole population cannot be taken down to the individual person
An ecological study cannot comment on the individual risk of outcome
Advantages and disadvantages of case-control studies
Advantages
- relatively cheap
- quick
- can be used to study a rare disease
- if the disease is rare, odds ratio can be used to estimate relative risk
- multiple exposures can be investigated
Disadvantages
- prone to bias (selection, recall, information)
- cannot be used for rare exposures
- cannot investigate multiple diseases
- concerns about reverse causality (outcome might actually influence the exposure)
- doesn’t measure incidence or prevalence
Explain how matching is done
- each case is matched to one or more controls (for e.g. age, sex)
- all the possible matches to the case in the control group are found and then 1 is selected at random
Advantages and disadvantages of cohort studies
Advantages
- evidence of exposure doesn’t depend on memory (no recall bias)
- exposure is measured before the disease onset so you can assume causality
- multiple outcomes for 1 exposure can be studied
Disadvantages
- slow
- expensive
- difficult to administer
- losses to follow up may introduce a selection bias (e.g. those that remain have a different health behaviour)
- exposure and diagnostic criteria may change over time
- outcome status may be influenced by the knowledge of exposure
What are the purposes of a systematic review?
- summarise and simplify vast literature
- increase the precision of estimates
- assess consistency and generalisability of findings
- explain inconsistencies
- highlight areas of certainty or deficiency of evidence
Steps for analysing a forest plot
- Sage overall what it shows
- Comment on the risk ratio
- Comment on the p value (<0.05 = can reject null hypothesis)
- Then talk about the different plots - study with most (biggest box) and least (smallest box) weight
- Talk about heterogeneity (I squared)
- If asked: comment on if confidence interval straddles null line (bad)
Describe fixed effect and random effects models - when and why you would use them
Fixed effects model
- used when there is homogeneity (low I squared)
- because we can assume that the studies are all estimating the same effect
- greater weight is given to larger studies (as smaller ones show sensational results)
Random effects model
- used when there is heterogeneity (high I squared)
- because it accounts for variability within and between the studies
- gives greater relative weight to smaller studies
Strengths and weaknesses of high risk strategies
Strengths
- can offer individual interventions that are appropriate
- can be effective and cost effective
- avoids interference with the low risk population
Weaknesses
- imperfect prediction of who is high risk
- medicalising of prevention and health - doesn’t address the broader social and environmental determinants of health
- limited efforts to interfere with the incidence of disease
- behaviourally inappropriate - singles people out who have to behave differently to everyone else (stigmatisation)
- ROSE PARADOX: e.g. high BP is associated with a high CHD risk, but most cases of CHD occur in moderate levels of BP that are commoner in the population. Public health should focus on small changes in everyone rather than only high risk individuals to make a true population-wide change
Strengths and weaknesses of population strategies
Strengths
- large benefit where risk is widespread
- behaviourally appropriate - don’t have to separate groups about
- no need for screening or identification
Weaknesses
- benefits for population may be large, but small for the individual
- depends on public and political motivation - usually low
- complex measures may be needed
- balance of benefit and harm may be narrow if you have to give something (e.g. adding fluoride to all water)
