Measuring Association Flashcards
define cause
any event, act, or condition preceding disease or illmess without which disease would not have occurred or would have occurred at a later time
what does disease result from?
the cumulative effects of multiple causes acting together (causal interaction)
what are the 3 types of causes?
- necessary cause
- contributing cause
- sufficient cause
descrine necessary cause
found in all case
describe contributing cause
needed in some case
describe sufficient cause
the constellation of necessary and contributing causes that make disease inevitable in an individual
what do you need before you can start determining causes?
a case definition!
what is causal complement/causal pie?
the set of factors that completes a sufficient causal mechanism
give an example of causal complement in tuberculosis
the necessary agent is mycobacterium tuberculosis, the causal complement is one’s susceptibility to the bacteria
define and give 7 examples of host factors
aspects of the host that contribute to a disease state
1. genetics
2. physiology
3. anatomy
4. behavior
5. occupation
6. constitutional
7. cultural
what is needed to determine relative risk?
one must examin each potential risk factor and calculate the relative risk
what is exposure?
exposure to the risk factor, not necessarily exposure to the causative agent
what is risk?
the proportion of individuals exposed to a risk factor that become ill
how is risk calculated?
the number of cases that become ill divided by the total number of people exposed to the risk factor
what does risk NOT tell you? how do you fill this knowledge gap?
does not tell anything about thosenot exposed to the risk factor who could also be ill; go talk to the people who are sick! shoe leather!
what is relative risk?
the relative chance of becoming a case upon exposure to the risk factor compared to those not exposed
what do you need to calculate relative risk?
case definition!!!
what does relative risk (risk ratio) compare?
compares the risk of becoming a case between the exposed and unexposed groups
how is risk ratio calculated?
risk in those exposed divided by risk in those not exposed
a/(a+b) over c/(c+d)
what is attack rate similar to?
very similaer to risk rate in both concept and calculation
what is attack rate commonly?
a measure of infectivity
what does attack rate do for a person-to-person spread infection?
limits the number of exposed individuals to those that are both exposed and susceptible
what would lower the attack rate?
vaccines
what is attack rate really a measure of?
communicability, or the ability of a disease to spread person-to-person
give the formula to calculate attack rate
(number of new cases among the population during the time period divided by population at risk) x100
what is calculating risk, relative risk, and attack rate all dependent upon?
knowing everything about the population that you have narrowly defined; true for both cases and controls where you know how many are involved and how many are susceptible
can the results of risk, relative risk, and attack rate be extrapolated outside of the defined population?
no
what must one do when the population is poorly defined or one cannot be confident that all of those exposed to the risk factor are accounted for?
play the odds to make statistical comparisons between two samples of peple, cases, and controls
how are controls selected when calculating odds?
controls should be selected from those most likely at risk for exposure, yet did not develop the disease
what does odds measure?
the odds of having been exposed to the risk factor versus the odds of not being exposed to the risk factor
what must odds be calculated for?
both the cases and the controls
how are odds for cases calculated?
number of cases exposed to RF over number of cases not exposed to RF
how are odds for controls calculated?
number of controls exposed to risk factor over number of controls not exposed to risk factor
what is odds ratio?
analogous to relative risk, but this calculates the odds for exposure among cases divided by those for controls
how is odds ratio calculated?
odds for cases over odds for controls
what is the only real difference between odds ratio and risk ratio?
the ambiguity of data on the population as a whole (more ambiguous with odds ratio)
what do the signs of odds ratio mean?
OR =1: no assocation between risk factor and disease
OR<1: negative association (risk factor protective against disease)
OR>1: positive association: risk factor indicative of disease
can odds ratio be extrapolated to other populations? what does it include?
yes, but includes a margin of error
why is there statistical uncertainty for each of the four groups in the contigency table made from odds ratio?
because we do not have that actual data for the entire population involved
what must you also calculate with odds ration due to the statistical uncertainty?
the error factor for the odds ratio
what is the error factor for the odds ratio?
the plus or minus; the confidence interval
when do you need to factor in the error factor of the odds ratio?
when extrapolating the odds ration to another population
how do you decrease the error factor of odds ratio?
increase the sample size
compare and contrast relative risk versus odds ratio
relative risk needs incidence of the disease, is prospective, and demonstrates temporality while odds ratio does not require incidence for its calculation and is retrospective
when is odds ratio a good estimate for risk rate?
in cases of low frequency disease
when can odds ratio overestimate the risk rate of a disease?
in a rare disease