Epi Final, I
What is the main objective of randomized clinical trials?
test the possible effect of a therapeutic or preventive intervention
What is the main difference between randomized clinical trials and experimental studies?
in experimental studies the investigator has direct control over study conditions
How is bias avoided in RCT?
animals randomly allocated to treatment and control groups
less opportunity for error since not controlled by investigator
What trials evaluate efficacy of therapeutic agent or novel procedure?
What evaluates whether agent or procedure reduces risk of disease developing in those currently free of disease? (ie. vaccination study)
What is the difference between open and blinded study? Double blinded?
: subject and investigator know which are assigned to tx
: investigator know but subjects are unaware of tx
: both subject and investigator are unaware
What are strengths of clinical trial?
researcher has control over study situation (not observational)
best measure for causal relationship
randomization minimizes confounding /removes selection bias
blinding eliminates selection and observation bias
What are weaknesses of clinical trial?
excessive control of study means it does not accurately represent real situation
large sample size needed/costly
long duration is new case occurrence is low
What type of testing is done to apparently healthy members to detect presence of disease?
what type of testing confirms or classifies disease status, guides treatment and provides prognosis?
Why are false positives a problem?
waste of money; subjected to more sophisticated testing
Why are false negatives a problem?
misdiagnosed during early stages when possibly still treatable
and allows infected individual to be source for others
What are 2 components of validity?
sensitivity and specificity
What is defined as proportion of subjects with disease that test positive (true positives)?
What is defined as proportion subjects without disease that test negative (true negatives)?
What is required to evaluate sensitivity and specificity of new test?
use gold standard test to evaluate before apply new test
Does the test have high sensitivity or specificity if you trust negative results, thus use the test to rule disease OUT?
(SNOUT and spin = sensitivity rules dz OUT)
Does the test have high sensitivity or specificity if you trust positive results, thus use test to rule disease IN?
(snout and SPIN = specificity rules dz IN)
What type of variable requires an established cut off point to determine if results are +/- ?
continuous variable (ie. glucose, blood pressure)
What type of variable gives a simple positive or negative result?
How do you calculate sensitivity?
true positives / total disease positives
When using the table, do you use columns or rows to evaluate sensitivity/specificity? What about for predictive value?
columns for sens/spec
rows for pred.value
How do you calculate specificity?
true negatives / total disease negative
What is positive predictive value? How do you calculate positive PV?
proportion of subjects with POSITIVE test results that HAVE the disease
true positive / test positive
What is a negative predictive value? How do you calculate it?
proportion of subjects with NEGATIVE test results which do NOT have disease
true negative / test negative
Epi Final, I
Epi Final, I