Midterm terminology Flashcards
What are Bradford hill’s criteria for causation?
9 criteria for causation that show that the subjects are causally related
Define strength of association
-strong associations provide firmer evidence of causality than do weak ones
-quantified by a measure (rate, ratio, etc.)
Note: the revers is not true– weak association do not disprove causation
Define consistency
-similar findings found in multiple studies (must be variable in methods, populations, and circumstances)
-consistency alone does not prove causation
Define specificity
-causal factor should lead to only one disease and that the disease should result from only this single cause
-when present it is a very powerful criterion
Define temporality
-exposure to the causal factor must PRECEDE the onset of disease
-this criterion is REQUIRED to establish causation
Define biological gradient
-an increase in the level, intensity, duration or total level of exposure to an agent leads to progressive increase in risk (ex. smoking and lung cancer)
-if a threshold exists below which no further harm is done, further reduction in exposure is unwarranted
Define plausibility
-the association needs to be plausible with known biological facts about the disease
-common sense and known biology must take precedence over statistical relationships
-biological plausibility is also contingent on the biological knowledge of the day
Define coherence
-available evidence concerning the natural history, biology, and epidemiology of the disease must “stick together” (cohere) to form a cohesive whole
-look at BIG picture = makes sense?
Define experimentation
Requires:
-experimental studies
-natural experiments
-in vitro lab experiments, and animal models
-comparative medical research
Define analogy
-implies a similarity between things that are otherwise different
-if one pharmaceutical drug (thalidomide) can cause birth defects, so might others
-weaker criterion but is often a starting place for further research
Describe a case control study
-retrospective
-observational
-looks at the outcomes then looks for exposures
Describe a cohort study
-prospective
-observational
-looks at exposures then watches for outcomes
Describe a cross-sectional study
-where an exposure and an outcome are measured at the same time
-ex. BMI and cholesterol level
Describe a randomized control trial
-prospective
-interventions study
-has an intervention group and a control group then looks for outcomes
Define reliability
how consistent the outcome of the tool is (does it give the same response each time the same person is tested)
Define validity
how well the tool measures what it is supposed to measure (is it actually measuring the correct thing, such as IQ)
What is the relationship between reliability and validity?
a tool can not be valid if it is not reliable
