Post-midterm 2 Flashcards
When to use odds ratio and rate ratio to calculate attributable risk %?
When frequency of the outcome is low (<0.10) the odds ratio may be used to estimate the risk ratio
What is ecologic fallacy
An ecologic fallacy occurs when one concludes incorrectly that an association found at the group or aggregate level also applies at the individual level.
When are ecological studies undertaken
- Only aggregate information on exposure is available
- The exposure of interest varies only at the population level
- Interested in both individual and group exposures (multi-level studies)
Limitations of ecological studies
Ecologic fallacy
Sources of exposure or outcome may not be accurate
Difficult to control for confounders, especially at individual level
May be impossible to establish temporal sequence
Strengths of ecological study
Quick and low cost
Generate hypotheses for further study at the individual level
Some associations may only be examined at the aggregate level
What is a cause
An event condition or characteristic that preceded the disease event and without which the disease event would not have occurred at all until some later time
What are extraneous variables
Variables that occur outside of the exposure-disease relationship
Examples: Confounding variables, effect modifying variables, neither
What is confounding
Confounding is an intermixing of the effect of an exposure with that of an independent risk factor for the outcome (disease), leading to an estimated association that no longer reflects the causal impact of the exposure of interest
What is mediation
Represents an intermediate step in the causal sequence between the exposure and outcome. Mediating variables are links in a chain of causal events connecting exposure to disease
Ways to control confounding in the design phase
◼ Restriction
◼ Randomization
◼ Matching
Ways to control confounding in the analysis phase
◼ Standardization
◼ Stratification
◼ Matching
◼ Regression (multivariable) models
What is restriction? what are the advantages and disadvantages
Eliminate variation in confounder (E.g., sex: recruit only male)
Advantages:
Effective, complete control of known confounder
Convenient, inexpensive, straightforward data analysis
Disadvantages:
Shrinks pool of available study subjects
Cannot evaluate primary relationship across confounder levels
May limit generalizability
What is randomization? Advantages and disadvantages?
Generate groups fairly comparable with respect to distribution of confounders in each exposure category
Advantages:
Control known/ unknown factors
Convenient, inexpensive, straightforward analysis
Disadvantages:
Usually applied to RCTs/ intervention studies
Works well only for large sample sizes
What is matching? Types? Advantages/Disadvantages?
Ensuring groups being compared are matches in terms of distribution of confounding variables. Type of partial restriction.
Pair matching: Each member of the control has to have the same value of confounding variable as the case group
Frequency matching: Frequency of cofounder is same in both groups
Advantages:
Easy, inexpensive, increase precision
Disadvantages:
Costly, time-consuming
Matching can no longer be evaluated as a risk factor
How is controlling confounding at analysis phase different
Must measure in study
Advantages of stratification and regression models
What is effect modification? How is it different from confounding?
Happens when an extraneous variable modifies the effect of the exposure of interest
Different because effect modification is not an artifact that can (or should) be adjusted away or controlled
It is a real aspect of the investigation
What is stratification and regression models?
Stratification: Divides samples into groups by levels of the confounding variable. Then adjusted for confoundings.
Regression: Produces adjusted estimates, can handle larger # of confounders, usually done with software
How do you know if uniform?
Visual inspection
Statistical test
What is primary prevention
Seeks to interrupt a destructive chain of causal events
For example: if people are becoming sick from drinking contaminated water, this is a modifiable link.
What is secondary prevention
Identifying disease in its earliest stages (screening)
What is tertiary prevention
Reducing negative impacts of established disease. Targets specific links in chains of causality that leads to better clinical outcomes
What are the 4 frameworks which inform epidemiological judgements/assessments of causation
Web of causation
Causal pies
DAGS
Bradford-Hill’s Criteria
What is web of causation
Population patterns of health and disease can be explained by a complex web of numerous interconnected risk and protective factors
T/F it is possible to interrupt the disease by cutting one singular factors from a chain of various factors
True
What is causal pie? Component/sufficient/necessary cause?
Proposes that sets of component causes combine to produce a causal mechanism.
Component: Single element of a constellation of components making up a sufficient cause
Sufficient: A cause that inevitably produces the effect; a minimum number of components causes which produces a disease (whole pie)
Necessary: A component cause which is a member of every sufficient cause for a disease
What is DAGS? How to create DAGS)
Visual representations of qualitative causal assumptions
Diagrams follow strict rules
◼ To create a DAG one must specify:
1) the causal question of interest, thus necessitating inclusion of exposure (E) and outcome of interest (D);
2) variables that might influence both E and D;
3) discrepancies between the ideal measures of the variables and measurements actually available to researchers;
4) selection factors that influence which individuals are represented in the study population; and
5) potential causal relationships among these variables (depicted as arrows connecting variables).
What is consistency
The association is observed repeatedly in different persons, places, times and circumstances
What’s temporality
Causal factor must precede the disease in time
The only one of Hill’s criteria everyone agrees with
Prospective studies do a better job establishing the correct temporal relationship between exposure and disease
What are some of Hill’s criteria
◼ strength of association
◼ consistency
◼ specificity
◼ temporality
◼ biological gradient
◼ biological plausibility
◼ coherence
◼ experimental evidence
◼ analogy
What is dose-response
A “dose-response” relationship between exposure and disease. Persons who have increasingly higher exposure levels have increasingly higher risks of disease.
What is biological plausibility
Based on current biological social models, the association does not conflict current knowledge. The association is plausible
What’s coherence
Similar to biological plausibility but engages a wider field beyond biology. Referring to the fit of an epidemiological estimate with results from any type of scientific research or theory
What is experimental evidence
Used to be criteria for weighing causality. Current epidemiology uses experimental evidence to indicate that a study has employed randomizations. Provides strong evidence for causation, but many epidemiologic studies are observational.
What is specificity
This criterion is rarely used. A given agent is always associated with only one disease and the agent can always be found with the disease.
Critical appraisal steps
◼ Step 1: Identify the study’s research question or hypothesis.
◼ Step 2: Identify the exposure and disease variables.
◼ Step 3: Identify the study design.
◼ Step 4: Assess for selection bias.
◼ Step 5: Assess for misclassification bias.
◼ Step 6: Assess for confounding and effect measure modification.
◼ Step 7: Assess the role of chance.
◼ Step 8: Address causality.
◼ Step 9: Assess generalizability.
◼ Step 10: Report the critical appraisal.
Importance of step 2: identifying the study design
Each design has strengths and weaknesses, identifying the design helps to organize critical thinking
Red flags of ecological, Cross-sectional, case-control, prospective cohort
What should a bias assessment specify beyond saying it has occurred?
Magnitude of bias, the direction of bias (over/underestimation)
When will selection bias occur in a case-control vs a cohort
Will occur in a case control if the probability of selecting cases and controls depends on exposure in some way that differs between cases and controls.
Occurs in cohort if the probability of selecting exposed and non-exposed depends on disease in some way that differs between exposed and nonexposed
two types of differential misclassification bias in case-control studies
Recall bias
Interviewer/observer bias
How to minimize differential misclassification
◼ Exposure status should be assessed in the same manner for both cases and controls.
◼ Blinding interviewers and participants if possible
◼ Use of objective markers of exposure
◼ Verify exposure in other ways
◼ Use of diseased control group
What is diagnostic suspicion bias
Type of differential misclassification bias.
Suspicion that the exposure is a risk factor could greater attention to detection of the disease in exposed subjects
Could increase sensitivity or decrease specificity for disease status is exposed
What is Berkson’s bias
Synonym for selection bias in hospital-based studies
When the combination of the study exposure and outcome increase the chance that exposed cases are admitted to the hospital results in artificially elevated odds of exposure in cases
What is selective survival bias
Occurs when prevalent disease cases are studied and if exposure is related to disease prognosis
What is protopathic bias
Occurs when early manifestations of disease cause a change in the pattern of exposures in cases
should confounding or effect modification be assessed first
effect modification
How should a critical appraisal be reported
Written up in a few pages or verbally in 5 to 10 minutes
