BIOL 437 Week Ten p.2 (Causality) Flashcards
direct causal association
- no intermediate factor
- more obvious
- elminating exposure will eliminate the adverse health outcome
ex. a trauma to the skin
indirect causal association
- involves one or more intervening factors
- often more complicated
ex. poor diet and stress may cause high BP, which can cause heart disease
factors of causation
- predisposing factors
- enabling factors
- precipitating factors
- reinforcing factors
predisposing factors
- factors already present that produce a susceptibility or diposition in the host to a disease or condition without actually causing it
ex. age, immune status, knowledge, beleifs
reinforcing factors
-help aggravate and perpetuate behaviours, disease, conditions, disability or death
positive reinforcing factors
- social support
- health education
- economic assistance
negative reinforcing factors
- negative peer influence
- poor economic conditions
enabling factors
- antecedants to behaviours, disease, conditions, disability or death
ex. services, living conditions, programs - can also be a result of lack of services or medical programs
precipitating factors
- essential to the development of diseases, conditions, injuries, disabilities and death
ex. infectious agent, lack of seat belt use, drinnking and driving, lack of helmet
evidence for causal relationship (Henle and Koch postulates)
- Organisms is always found with the disease
- Organism is not found in those without the disease
- Organism when isolated from one who has the disease and is cultured, produces the disease
direct observation
-lot of knowledge gained in this manner
inference
-essential to scientific acitivty
>not everything can be observed
latency
ex. long period between exposure to cigarettes and onset of lung cancer
induction
Example:
- 2-week long induction period of measles and its infectiousness: barrier to understanding transmission
- confoudned seasonality link
Test of operational hypothesis
- design study and collect data
- analyze data and make conclusions about study hypothesis
- modify hypotheses if needed
type 1 error
-when hypotheis is rejected but is actually true
type 2 error
-when hypothesis is not rejected but is actually false
falsification of a hypothesis (Karl popper)
-more informative than corroboration of a hypothesis
>a single counter example forces modification
>studies should attempt to refute rather than confirm hypothesis
-problematic in epidemiology
Karl Popper falsification process
- requres a large body of knowledge
- negative result often cannot refute orignail hypothesis due to many sources of bias that can mask underlying associations
- makes it hard to refute an epidemiolgoical hypothesis which limitis the application of Popper’s model
common sense
- alternative model of scientific progress
- judgement under uncertainty that characterizes human experts and computer-based expert systems
- a blatant violation of of logic theory but is often how we actually operate
what is causal inference?
- a conclusion about the presence of a health-realted event and reasons for its existence
- provide scientific basis for medical and public health action
- made with methods comprising lists of criteria applied to the results of scientific studies
what is statistical inference?
- draws a conclusion about a population based on information from sampled data
- probability is used to indicate level of reliability
- possibility that chance, bias or confounding explain a statistical association should always be considered
statistical inference
-used in evaluating the data for use in causal inference
possilbe non-causal reasons for an association
- characteristic of study group
- problems with measuring disease or exposure
- another factor affecting disease and putative cause
causal inference overall
-attempt to eliminate all possible non-causal reasons for an observed association
causal inference is a reasoning process based on
- what is believed to be true
- the previaling concepts of disease
- knowledge at the time
- erroneous and potentially ignorant beleifs
Viral diseases
-do not follow Koch’s postulates
1. disease production may need co-factors
2. Cannot be cultured as they need living cells to grow
3. Pathogenic virusus can be prsent without clinical disease
>subclinical infections, carrier states
true association?
-if controls have been selected in a mannor that favours non-exposure then the association of exposure to disease may be false, due to study design alone
if it is real, is it causal?
-is there a confounding factor that links the exposure to the disease?
>YES=then it is not a causal association
3 methods of hypthesis formulation in disease etiology
- method of difference
- method of agreement
- method of concomitant variation
method of difference
-frequency of disease occurence is extremely different under different situations or conditions
-if a risk factor can be identified in one condition and not in a second, it may be that factor
>or the absence of it that causes the disease
method of agreement
- if risk factors are common to a variety of different circumstances
- risk factors have been positively associated with a disease
- probability of that factor being the cause is extremely high
method of concomitant variation
-the frequency or strength of risk factor varies with the frequency of the disease or condition
>increased numbers of children not immunized causes the incidence rate to go up (Ex. measles)
strength of association
- excess of disease associated with exposure
- magnitude of ratio of incidence in exposed to incidence in the unexposed
consistency
-assocation has been repeatedly observed:
>by different people
>in different places
>in different circumstances
>at different times
-guards against associations due to error or artifact
-not necessarily free of bias
specificity
- relationship between exposure and disease is specific in various ways
- specific types of exposures are more effective
- one agent can still contribute to multiple diseases
temporality
-first expsoure, then disease
-not always possibile to document sequence
>long lag periods
>subclinical disease
biological gradient
- dose-response relationship consistent with the hypothesized conceptual model
- incremental change is disease rates in conjunction with corresponding changes in exposure
- consider threshold and saturation effects
plausibility
- does it make biological sense?
- common sense
coherence
- does it fit with known facts
- include knowlege about distributions of exposure and disease and lab experiment results
experimental evidence
-some designs give more convincing evidence
-intervention studies can give the strongest support
-demonstrating under controlled conditions that changing the exposure causes a change in the outcome
>not always possible though, especially when dealing with humans
analogy
-accept causal inferences that resemble others already accepted
the role of chance
-characteritics of subjects may vary from sample to sample
>as a result, an association between an exposure and outcome may be from chance
-sample size directly related to chance
-increase sample size to decrease chance
web of causation
-complex sets of events or conditions caused by many activiites connected to a common core or experince or event
search for cause vs decision-making
- all scientific work is incomplete wheter it is observational or experimental
- sometimes we have to work with what we know in the moment
