Test Validation Flashcards
(56 cards)
What is the statistical parameters which allows you to have a rationale for judging the effectiveness of a particular diagnostic test, such as a sonographic examination, referred to as?
(Trying to determine if a test can be utilized in place of the Gold Standard)
Test validation
In order for a test to be effective, it must either (blank) or (blank)
- Selectively identify patients with the specific disease
or
- Selectively rule out patients as not having the specific disease (like a screening test)
List the three parameters commonly used to judge the efficiency of a diagnostic test:
Sensitivity, specificity, and accuracy
What is the standard that all tests are compared to, and definitely tells physicians whether or not a patient has a specific ailment?
Gold Standard
List an example of the gold standard being used:
Histology from a biopsy indicating a cancerous mass
What is used to when there is not a gold standard to compare to?
If there is no gold standard test, then either
- An autopsy
- clinical information/findings during surgery are used
What is the gold standard for a carotid artery stenosis (CAS)?
Digital subtraction angiography (DSA)
In what scenario would u/s not be able to detect a CAS, resorting to the gold standard?
u/s may or detect CAS in cases of too much shadowing from plaque (can’t see narrowing)
In what scenario would u/s not be able to detect a DVT, resorting to the gold standard?
U/s may not detect blood clots in situations where patient has too much fluid in the body (edema) to penetrate via u/s
What is the gold standard for a DVT?
Venography/venogram
What is the term for when the testing procedures demonstrates disease and the disease is really present?
True Positive (TP)
(Test and GS agree there is pathology)
List an example of a “true positive” finding:
Both a BLEVDOPP u/s and a venogram agree there is pathology
What is the term for when the testing procedures demonstrates no disease, and no disease is present.
True Negative (TN)
Both the test and GS agree there is no pathology.
What is the term for when the testing procedures demonstrates disease that is not really present?
I.e. the test said abnormal (pathology), but the gold standard said normal (no pathology)
False positive
Test being evaluated is falsely positive
List an example of a false positive
VDOPP u/s indicated DVT, but Venogram (GS) was normal.
Thus, there wasn’t really any pathology.
Term for when the testing procedures demonstrates the presence of no disease, and disease is actually present.
I.e. The test said normal (no pathology), but the gold standard said abnormal (pathology)
False negative
Test being evaluated is falsely negative
Which parameter is a measurement of how real a disease finding by that exam is?
Specificity
Which parameter is a measurement of how often a test will miss real disease?
Sensitivity
(Positive)
What is the ability of a test to detect a disease?
Sensitivity
What is the ability of a test to identify normality (rule out pathology)?
Specificity
What is the formula for sensitivity?
True positives (TP) divided by true positives (TP) plus false negatives (FN)
T/F
Low sensitivity is acceptable if known.
False
Low sensitivity is dangerous; patients who are positive may not receive the treatment they need
What is the term for how much you can trust a test?
Specificity
Minimizing false-negatives will increase the (blank)
Sensitivity