Diagnostics and Lab Testing Flashcards
What is the definition of sensitivity?
The ability of a test to correctly identify those with the disease (true positive rate)
What is the equation for sensitivity ?
True positives / (true positives +false negatives)
What is the definition of specificity?
The ability of a test to identify those without the disease (true negative rate)
What is the equation for specificity?
True negatives / (false positives + true negatives)
What is the positive predictive value?
The disease percentage in those testing positive
What is the equation for positive predictive value?
True positives/ (true positives + false positives)
What is the negative predictive value?
The percent testing negative in those without disease
What is the equation for negative predictive value?
True negatives/ (true negatives + false negatives)
What is the likelihood ratio?
The likelihood of a result meaning disease versus not meaning disease
They are basically used to assess the value of performing a certain test, and predicting whether it will significantly change the pre test probability.
These can be useful to figure out whether to perform a certain test, when put in combo with pre test probability
What is the equation for positive likelihood ratio?
Sensitivity (1-specificity)
What is the equation for negative likelihood ratio?
(1-sensitivity)/specificity
What do you use a receiver operating curve for?
To see the ‘trade off’ between sensitivity and specificity, to allow optimisation of a cut off point for a test
What is the change in probability if the likelihood ratio is:
a) 2
b) 5
c) 10
d) 0.1
e) 0.2
f) 0.5
a) 15
b) 30
c) 45
d) -45
e) -30
f) -15
ie. the higher positive value, the more likely the post test probability to change = DO THE TEST
ie. the higher the decimal point (closer to 1), the less likely the post test probability to change = DONT DO THE TEST
What is analyte specificity?
The ability of a test to detect the analyte alone. This can be interfered with through:
- Related molecules
- Matrix effects
- Antibodies etc
What is the limit of detection?
The smallest conc of an analyte that can be distinguished from zero
What is the limit of quantification?
The smallest conc of analyse that can be measured with ACCEPTABLE PRECISION (SD>20%)
What is precision?
The repeatability of a measurement (measure of random error)
What is accuracy?
How close the result is to the actual amount present (measure of systematic error)
What is spectrophotometry?
A lab technique in which light is passed through a sample, and the intensity correlated with absorption over certain wavelengths
What are antibody based methods?
Immunoassays containing antibodies that bind to a specific part of a foreign antigen, called the epitope. They can be POLYCLONAL or MONOCLONAL.
Usually they are attached to an enzyme that produces a colour or a flurophore
What is the difference between polyclonal and monoclonal antibodies?
Polyclonal - high affinity, recognises multiple epitopes
Monoclonal - high specificity, very expensive
What are the two types of immunoassay?
Sandwich - signal produced is proportional to the amount of analyte present
Competitive - signal produced is inversely proportional to amount of analyte present (as it accounts for whatever isn’t captured by the antibody)
What are heterophilic antibodies?
Abs produced against poorly defined antigens, which are generally weak with multi specific activities. They can sometimes interfere with immunoassay tests
What are immunoassays used for?
Primarily diagnosis of primary malignant tumours, such as HER2 staining in breast cancer
What are the limitations of immunoassays?
- Cross-linking
- Background staining
- Difficult to standardise
What is 2d electrophoresis?
A lab technique which separates out analytes based on structure and weight
What is mass spectrometry?
A lab technique used to target detection of specific molecular fragmentation. It sorts ions based on their mass to charge ratio.
What is Bayes Theorem?
It describes the probability of an event, based on prior knowledge of conditions that might be related to the event.
ie. it considers the test characteristics (likelihood ratio), as well as the pre-test probability
What should be the outcome if you have
a) very low pre test probability
b) intermediate pre test probability
c) very high pre test probability
a) consider if you really need to test and treat?
b) TEST and treat on the basis of the results
c) consider if you really need to test, or should just get on with treatment?
Why is it useful to do a test if the pre test probability of having a disease is intermediate?
It influences interpretation of a borderline result
