Rod Jackson Flashcards
What is the study if epidemiology
Frequency or occurence of disease in populations
Why do epidemiologists measure the frequency of health + disease in different populatuions
Find out the causes of poor health and how to improve it
What is the goal of epidemiology
Measure the frequency of health and disease in different populations to find out the causes of poor health and how to improve it.
What is the numerator
Cases of disease
What is the denominator
The population
Why do we measure the frequency of disease in different populations?
If frq of disease is different between two populations (or in same population at different times, can help identify the causes
What is age standardisation
Adjust or standardise the age structures of each population so they can be meaningfully compared, and take into account the differences between the denominators.
What is the equation of epidemiology
Epidemiology = N/D
What are EGO and CGO
The occurence of risk of dis-ease in populations
Equation for EGO and CGO
EGO = a/EG
CGO = b/CG
What are the design features of a cohort study
- start with a group pf participants
- allocate them into at least two groups by measuring things
- follow over time
What is incidence
If the number of dis-ease events that occur are counted forward from a starting point, over a period of time
What is prevalence
If the number of people with dis-ease are counted at one point in time, then T = 1
Does incidence or prevalence have a numerical outcome
Incidence
What are the two ways to define prevalence
- Categorical measure (categories): Prevalence of fast heart rates/people in population
- Numeral measure (numbers): average heart race = sum of every persons heart rate/people in population
What study allows incidence and prevalence to be measured at the same time
Cross-sectional
What are the different ways to measure numerator when measuring prevalence
If its difficult to measure when a disease event occurs or it occurs frequently, often measure if (yes/no) it has occured during a prior period but not the total occurences.
What are the strengths of incidence
- Determined by only dis-ease risk in population (clean measures)
- includes events (N) population (D) and time (T)
What is the weakness of using incidence
Difficult to observe events over time
What is the strength of prevalence
Relatively easy to measure as you ‘stop time’ and count
What are the weaknesses of prevalence
- Determined only by incidence, cure and death rate (dirty measure)
- includes only events (N) and population (D) - less info
What are RCT’s
Like cohort studies, but participants are randomly allocated to EG and CG.
What is the best study to answer effects of treatments
RCTs but only if its ethical and practical
How to calculate risk difference
EGO - CGO
What happens if RD = CGO
Risk difference = 0
How to calculate risk ratio
EGO/CGO
What happens if EGO = CGO
Risk difference = 0
What happnes if RR < 1
This is relative risk reduction (RRR)
How to calculate relative risk reduction
Calculated by subtracting relative risk from 1.0 and multipluing by 100
What happens if RR >1
This is relative risk increase
How is relative risk calcualated
By subtracting from 1.0 from RR and multiplying by 100
What are the differences between RR and RD
- RD has units but RR does not
- RD also called absolute risk differene, RR also called relative risk
Strength and weakness of RR and RD
- Most people understand RR better than RD but relative measures of risks and benefits are deceptive and provide less information than absolute risk differences
- all decisions should be based on RD
Where does error occur
Wrong people are recruited into a study (GATE) or the right people are put in the wrong GATE frame category.
What are non-random error due to
- poor study design
- poor study process
- poor study measurements
What is study validity
Study with only a small amount of random or non-random eorror considered valid study
Randomised control studies
- Cohort study except how to get into CG and EG
- allocated randomly
- called experimental
- always longitudinal
Whats the main problem with allocation
Confounding: when exposure mixed with another factor that is also associated with the outcome, has ‘confounding’ error. A potential problem in all observational studies
How to deal with confoundment
Divide ‘stratify’ the study into sub-studies so participants with the confounders are all in one sub-stratified analysis
What does adjustment do when dealing with confoundment
Adjustment involves dividing the participants into strata and comparing heavy and moderate drinking in each strata. This reduces the effect of the confounders.
Where does non random error occur (4)
- Recruitment errors
- allocation errors cause confounding
- maintenance errors
- (BO) measurement of exposures and outcome errors
The differences of allocations in ecological studies and individual studies
Ecological: allocated populations to EG and CG.
Individual: allocate individuals to EG and CG
Is confounding common in ecological studies, and explain why
Yes, because there are a lot of differences between countries and it is difficult to measure these differenes and adjust for them
Does an ecological rct increase or reduce chance of confounding
Reduce
What does it mean by measuring the exact ‘truth’
The study of participants are moving targets, so….
Identical measurements of exposures and outcomes in the same or similar people can change from moment to moment
What is random measurement error
Identical measurements of exposures and outcomes in the same or similar people can change from moment to moment
Major causes of random errors
- only using a sample of all the people in population
- even identically designed and implemented studies include participants with identical characteristics, so could still produce different studies
- these differences are random errors
What does a smaller sample mean
Smaller the sample, the greater the chance the sample will be different from the whole population
What is the 95% interval
A measure of the amount of random error in our estimates of EGO and CGO, RR and RD in the whole population, when you have only done one study
What is the 95% interval
A measure of the amount of random error in our estimates of EGO and CGO, RR and RD in the whole population, when you have only done one study
95% confidence interval defintion
There is about a 95% chance that the true value in a population lies within the 95% confidence interval
In 100 studies using samples from the same population, 95/100 of the 95% CI;s will include the true value for the population
Confidence interval describes…
the range of results likely to include the true result in the population
What study design measures incidence and prevalence at the same time
Cross-sectional
What is the most useful study application of an ecological study
Investigating international trends and causes of dis-ease incidence and prevalence
True or false, an ecological study can be either an RCT, CS or a CSS
True
What is the most useful application of a cohort study
Investigating the caues of dis-ease incidence
What is the most useful application of a cross-sectional study
Measuring the prevalence of disease in different groups and population
Main strength of an RCT
Randomisation minimises confounding
Main strength of a cohort study
- Few ethical issues
- Cheaper than RCT’s
- Measured before outcome so avoid recall bias with clear time sequence between exposure and dis-ease outcomes
- Participants more likely a represensative of general populations
Main strength of a cross-sectional study
- Cheaper and completed quickly than RCT’s or cohort studies
- Best desing for assessing prevalence of disease in a population
- No maintenance error as no follow-up
Main strengths of ecological studies
- Usually cheaper and quicker than all other studies because uses already collected data
- Random error low
When is it useful to conduct an ecological study
- When the majority of populations are exposed but others are not.
- Efficient for rare outcomes
Main weaknesses of a RCT
- Ethical limitations
- Logistically difficult: follow up difficult and costly
- Usually small as large studies expensive
- Participants often not represensative of general populations
- Maintenace error common
Main weaknesses of a cohort study
- Confounding common
- Maintenance error common because exposure not controlled by investigators
Main weaknesses of a cross-sectional study
- Uncertain time sequence, limits interpretation of cause and effect
- Confounding common
Main weaknesses of an ecological study
- Confouding common
- Maintenance error common
Difference between the three study types
RCT randomised randomly, while the other two allocated by measurement.
RCT and Cohort both have vertical time arrow going down, cross sectional has a horizontal time arrow going right to left
A similarity individual participant studies
Every person in P is identified and individually allocated to EG or CG
What studies meausure incidence, prevalence, or both
- RCT’s and cohort studies measure both icnidene and prevalence
- Cross sectional only measures prevalence
Which study types combine multiple stduies in a meta-analysis
RCT’s and cohort (only when sample size is too small)
Which study types combine multiple stduies in a meta-analysis
RCT’s and cohort (only when sample size is too small)
Why are cross-sectional studies never combined using systematic reviews and meta-analyses
- Cross-sectional: prevalence studies are not usually combined because it does not make sense to estimate aerage prevalence
- Ecological: not combined because generally very large with low random error
