GATE Frame Flashcards
Numerator
Number of people from study population in whom dis-ease occurs
Denominator
Number of people in a study
Population
Any group of people who share a common factor
Quantified
Data that can be counted
Two categories of quantified data?
Categorical
Numerical
Non-observable event
Can’t easily observe so you measure at a point in time
Event
Counting as it occurs (eg road traffic accidents)
Occurrence
The transition from a non-diseases state to a diseased state
GATE
Graphic approach to epidemiology
What is dis-ease?
Any health-related event (death/ill health easier to measure than well-being)
What is epidemiology?
The study of how much disease
Occurrence of disease calculation?
Numerator/denominator
Explain gate diagram
Triangle= participants Circle= EG/CG Square= numerators (disease outcomes) Time= vertical and horizontal
Exposure group occurrence equation
EGO = a/EG
Comparison group occurrence equation?
B/CG
Can numerical measures be recorded as categories?
Yes
Example of categorical measures?
High vs low salt intake
Examples of numerical measures
Average (mean)/ EG
Incidence?
Number of onsets of dis-ease occurring during a period of time (eg raindrops)
Can incidence be observed as it occurs?
Yes
What type of quantified measure is incidence?
Categorical
How is incidence normally written?
As a percentage of people with the disease in a specific period
Equation of incidence
A/EG during time
Example of high incidence, low prevalence?
Flu
Example of low incidence, high prevalence
Obesity
Prevalence?
The number of people with a disease at one point in time
When is prevalence used?
When the transition from non-diseases to diseased is not easily observable
How do people leave the prevalence pool?
By death or cure
What does prevalence assess?
The amount of disease (burden) at a point in time
What is prevalence not useful for?
Identifying the cause
Prevalence calculation?
A/EG at a point in time
Two types of prevalence?
Point (at one point)
Period (in a set time frame looking back)
Calculations to compare disease between groups?
Risk ratio
Risk difference
Relative risk?
Risk ratio (ratio of occurrence)
Relative risk/risk ratio/RR equation?
EGO/CGO
Absolute risk?
Risk difference (difference between occurrences)
Risk difference equation?
EGO-CGO with units
No effect value RR?
1
No effect value RD?
0
Relative risk increase/reduction?
+- from 1 x 100
Odds ratio?
A/b /c/d
When is odds ratio more appropriate?
When disease becomes more common as two relative estimates increase
Types of error?
Random
Non-random
Other names for non-random error
Bias
Systematic errors
Validity problems
Define RAMBOMAN
Recruitment Allocation Maintenance Blind Objective Measurement Analysis
Two types of recruitment errors?
External validity error: study findings are not applicable
Selection bias
Example of allocation error?
Confounding
What study avoids confounding?
RCTs
What do analyses involve?
Adjusting for confounding etc, with stratified analysis, and checking risk analyses
What causes random error?
Chance
Three types of random error
Random sampling error (will never be 100% representative)
Random measurement study (our ability to measure factors in the same way is subject to difference
Randomness in human nature
How do we estimate random error?
With a 95% confidence interval
Definition of a 95% confidence interval?
A range of values of a particular measure derived from a single study that is likely to include the true value in the underlying population
What does a wider CI mean?
More random error
Will a 99% or 95% CI have a wider interval?
99%
How would you write a 95% CI?
There is a 95% probability that the true value of EGO in the whole population from which the study participants were recruited lies between 8 and 10
If EGO and CGO confidence intervals do not overlap we call it?
Statistical significance
What happens is the CI for RR or RD crosses the no effect line?
There’s too much random error to determine if there’s a real difference between EGO and CGO
What does width of CI depend on?
Number of events in the study
What is meta analysis?
Combining studies to generate a summary estimate of effect (and alternative to a large study)
What type of study often uses meta analysis?
RTCs
Study objective of RCTs?
effects of different interventions (exposure) on disease incidence in different groups
Main application of RCTs?
Studying the effect of interventions (ie new therapies)
Main design features of RCTs?
Longitudinal
Experimental
Participants randomly allocated to either study exposure or comparison exposure and dis-ease outcomes measured during a follow-up period
Main strength of RCTs
Randomisation minimises confounding
Main weaknesses of RCTs
Ethical limitations
Logistically difficult, long term follow up difficult and costly
Large studies expensive so usually too small (random error is an issue)
Maintenance error common
Study objective of cohort studies
To investigate associations (effects) between risk/prognostic factors (exposures) and disease incidence in different groups of individuals
Main application of cohort studies
Studying cause of dis-ease incidence or the effects of interventions
Main design features of a cohort study
Longitudinal
Observational (non-experimental)
Participants allocated to exposure and comparison by measurement and disease outcomes measured during follow up
Main strength of cohort study
Cheaper than RCTs
Exposure measured before outcome
Avoids recall bias
Provides clear time sequence between exposure and disease outcome
Main weakness of cohort study
Confounding
Maintenance error common in long term situation
usual objective of cross sectional studies
To measure disease prevalence in defined populations. To investigate associations between exposure and disease prevalence
Main application of cross sectional studies
Measuring burden of disease in different populations
Main design features of cross sectional study
Observational
Participants located to EG/CG by measurement and outcome is measured at the same time
Main strengths of cross sectional study
Cheap
Quick
Best for assessing burden/prevalence of a population
No maintenance error as no follow up
Main weaknesses of cross sectional study?
Uncertain time sequence (possible reverse causality) limits interpretation of cause and effect
Confounding common
Study objective of ecological study
To investigate associations between exposures and disease prevalence or incidence in different populations
Main application ecological study
Studying the causes of disease incidence and prevalence
Main design feature of ecological study
Longitudinal or cross-sectional
Non-experimental or experimental
Exposure and comparison allocated to groups rather than individuals
Main strengths of ecological studies
Cheap and quick
Useful when majority of some populations are exposed and others aren’t
Efficient for rare outcomes
Main weakness of ecological studies
Confounding very common
