Epideomology Flashcards
incidence
the rate at which new events occur in population
should not include cases that occured or were diagonosed eaelier
prevalance rate
proportion of people in pop who have the ds in a sp period /point of time
diff bw period and point prevalance
burden of ds
whether an outbreak is occuring or not
pre test probabilities
sensitivity
specificity
post test probabilities
positive and negative predictive values
1-sensitivity=
false negative rate
accuracy
true pos and neg/total screened patients
selection bias
random,independent sample
samole not representative
berkson bias type of
selection/sampling bias
measurment bias
control group
lead time bias
use life expectancy
experimenter expectancy/
double blind:not to tell who is control and who is main
design bias
parts of study do not fit together
random assignment
confounding bias
unanticipated factors obscure results
combine results from multiple studies
metaanylasis
recall bias
subjects cannot remember accurately
confirmatiin
late look bias
severly ds individuals are not uncovered
stratify by ds severity
pygmylion effect is
experimentor expectancy bias
observational studies
case report case series cross sectional case control cohort
cohort
risk factors vs non risk factors
prospective search for who gets the ds
(tell I,C not P)
case report
n=1
case series
n>1
features of ds can be ruled out
but no control
cross sectional
tell about only prevalance
case control
ds vs non ds
retrospective search of risk factors
tell C,not I,P
analysing cohort study
RR
AR
for analysing cross sevmctiinal studies
chi square
phase 1
safety in healthy volunteers
phase 3
efficacy and side effects
phase 2
protocol and dose level,how to administer
control groups
source of comparison to be certain that the experiment group is being affected by the intervention and not by other factors
- placebo
- std of care
definitive trial
3
double blind RCT
neither the subjects nor the researchers who is getting tt/placebo
cross over study
all subjects receive intervention but at different times
how ll u assess data from following observational studies
- cross sectional studies
- case control
- cohort
- chi square(ass of risk factor and ds)
- odds ratio(many risk factors for single ds)
- RR(single risk factors aff many ds)
what does odds ratio tell?
how strong is the risk factor
interpretation of NNT and NNH
- how many people do u have to treat to stop one ds
2. how many people do u have to expose to get one case
Berksons bias
Selection or sampling bias created by selecting hospitalised patients
Pygmalion effect
Describes researchers belief in the efficacy of treatment that can potentially effect the outcome
Experimental expectancy
