Diagnostic Tests Flashcards
diagnosis
classify animals as having a disease (or other health related state) or not
uses of diagnostic tests
clinical medicine: treatment and prognosis
surveillance: identify changes in disease status
international trade: keep infected animals out
research: identify and monitor research subjects
Dichotomous tests
only 2 possible answers
positive or negative
e.g radiographs, MRI, PM
continuous tests
continuum of possible answers, still need to classify as positive or negative
e.g serum chemistry, antibody titers, CBC
hematology diagnostic tests
measures amount of different cell types in a venous blood sample
biochemical diagnostic tests
measures enzymes, metabolites, proteins, etc-usually in venous blood sample
immunological diagnostic tests
use antigen to measure antibodies or vice versa
example of immunologic tests
SNAP FIV/FeLV
pathogen detection tests
detects the pathogen itself (or specific parts of it)
ex: microscopy, culture, virus isolation, PCR
T/F some test for epidemiological investigations “fingerprint” pathogens or discriminate infected from vaccinated animals (DIVA)
true
T/F usually cant measure the disease itself
true
what is measured?
‘something’ (chemical, antigen, etc) that is present in a certain quantity when an animal has a certain disease/underlying pathogen
test value
how do diagnostic tests work?
sample taken –> test is run and test value is produced –> decision i (+/-) is made by test or diagnostician (test result)
cut off value
determine experimentally as the value that minimizes false positive/ false negative results
T/F usually there is a clear separation in the test values between diseased and non diseased
false
T/F some non-diseased animals may have higher test values than some of the disease and vice versa
true
T/F can have false positives/negatives
true
True positive
diseased animal that tests positive
true negative
non-diseased animal that tests negative
false negative
diseased animal that test negative
false positive
non-diseased animal that tests positive
how are gold standard tests used
true disease status of individuals is often determined using a gold standard test, against which the performance of a new test is evaluated
disadvantages of gold standard test
labor intensive, impractical, highly invasive, slow and/or expensive
every test is evaluated by how accurately it classifies:
diseased individuals as test positive
non-diseased individuals as test negative
how to evaluate a test
test group of disease animals
test a group of non-diseased animals
establish a cut-off value
what are the 2 parameters to evaluate diagnostic test
sensitivity (Se)
Specificity (Sp)
diseased animals are used to determine
sensitivity (of the new test)
non diseased animals are used to determine
specificity (of the new test)
T/F Se and Sp are proportions
true
determining sensitivity
test a group of diseased animals
proportion of diseased animals that the test correctly classifies as positive
what are some factors that can cause sensitivity to be low?
few/small amounts of whatever the test measures
samples are degraded before testing
T/F a sensitivity of 100% has no false negatives
true
classifies all diseased animals as positive
In a low Se test (75%)
25% of diseased not detected and are false negatives
T/F Se doesn’t tell us anything about how well the test performs on non-diseased animals
true
determining Sp
test a group of non-diseased animals
proportion of non-diseased animals that the test correctly classifies as negative
what are some factors that can cause specificity to be low?
cross-reactivity (mistakenly identified as another)
samples are contaminated
T/F high SP (100%) has no false positives
true
correctly classifies all non-diseased animals as negative
in a low Sp test (75%)
25% of non- diseased animals classified as positive (false positive)
T/F Sp doesnt tell you anything about how well the test performs on diseased animals
true
what is a balanced test
Equal Se and Sp
equally misclassifies diseased and non-diseased
SnNOut
sensitive test when Negative rules disease Out
correctly classifies ALL diseased animals as POSITIVE
SpPIn
Specific test when Positive rules disease in
remember-looks at non-diseased animals so if positive can’t rule it out
T/F test with 100% sensitivity will have no false negatives but may have false positives if Sp is low
true
any negative results are from non-diseased animals
T/F test with 100% specificity will have no false positive results but may have false negatives with low Se
true
any positive results are from diseased animals (rules it in)
maximize sensitivity when:
need to detect all diseased or infected animals
maximize specificity when:
cost of false positive is high and dont care if there are lot of false negatives
ex: need to cull positive animals
how to use Sp and Se in series
1st test: high Se
- detects most/ all of diseased animals in population
- negatives are true negatives
2nd test: high Sp
- test all that tested positive on the first test
- all test positives from second test are true positives (eliminates false positives from the first test)
T/F Se and Sp tell you the probability that an animals which tests positive is truly diseased
false
tells you the probability of a diseased animal testing positive or non-diseased testing negative
PPV
positive predictive value
true positive/test positives
NPV
negative predictive value
true negative/test negatives
how to calculate Se
true positives/ (true positives = false neagtives)
how to calculate Sp
true negatives/ (true negatives + false positives)