Lecture 21+22+DLA Flashcards
(37 cards)
meta analysis
the combination of many studies to combine results
can obtain a more precise estimate of effect size
Which of the following is an advantage of the random effects model over the fixed effect model?
It gives more realistic results when there is heterogeneity among studies
Temporality
the exposure occurred before the outcome
the exposure could be due to the outcome
Strength of Association
measured by risk ratio and rate ratio
A strong association is more likely to be causal
A strong association is unlikely to be due to
undetected confounding or bias
Consistency
An association found in several different studies
with different study designs carried out under
different circumstances with different populations
more likely to be causal
less likely to have bias
Reversibility
Experimental evidence to show that the removal of the exposure leads to a reduction in the risk of the outcome
Provides very strong evidence in favor of a
causal relationship
Multi-causality
The multi-causal paradigm is the dominant theory of causation in contemporary epidemiology
Random Error/ Chance
Random error reflects a problem of precision in assessing a given exposure-disease relationship and can be reduced by increasing the sample size
affects internal validity
Systematic Error/ Bias
reflects a problem of validity of the study
selection bias:
Methods used by the investigator when recruiting
individuals for the study, from factors affecting the
study participation
information bias:
Systematic distortions when collecting information
about exposures and outcomes
recall bias
example of information bias
the ability to recall past exposure is dependent on case or control status
ex: inaccuracy of memory / peer interference
response bias
anything in a survey that would influence a response
can be question bias or interviewer bias
Ex: lack of privacy while doing survey
Berkson’s bias
a form of selection bias that impacts hospital based studies
the hospital population is atypical compared to the normal population
healthy worker effect
found in work-force studies
results in better health status than the general population
hawthorne effect
the subject will change their behavior due to the fact they are being observed
lead time bias
seems to improve survival time, but just increases disease time
use mortality instead of survival rate
length time bias
outcome appears better in screened group because more have a better prognosis
count all outcomes
compliance bias
outcome in screened group appear better just due to compliance, not screening
count all outcomes
confounding bias
a systematic distortion in the measure of an association is distorted
confusion of effects
B-cells: what type of antigen
protein, carb, lipid, NA
T-cells: what type of antigen
peptide fragments (protein)
antigen must first be processed by APC (dendritic cell)
How do antigen and lymphocytes meet?
secondary lymphoid tissue
MHC
cell surface molecule that is required for:
antigen processing
antigen presentation
graft rejection
found in chromosome 6; HLA gene
does not change
class I MHC
expressed on the majority of nucleated cells
alpha chain: transmembrane (a,b,c)
beta chain: Beta 2 microglobulin
MHC 1 binding
alpha 3 binds to CD8 on cytotoxic T cells
presents endogenously
alpha 1 and 2 are highly polymorphic
