Flashcards in 06. Diagnostic Testing Deck (18):
"True positive rate"; Proportion of people with disease who have a true positive test result; How good a test is at detecting people with disease; Insensitive to prevalence
"True negative rate"; Reflects how good the test is at IDing people without disease; unlikely to produce false positives; insensitive to prevalence
Test positive & disease present
Test positive & disease absent
Test negative & disease absent
Test negative & disease present
Likelihood ratio (LR) definition
The likelihood that a given test result would be expected in a patient with the disease compared to the likelihood that the same result would be expected in a patient without the disease. Can be a positive or negative value.
Importance of likelihood ratio
Useful for assessing the value of performing a diagnostic test. Use the sensitivity & specificity of the test to determine whether the test result will significantly change the probability that a disease exists.
[Sensitivity/ (1-specificity)]; Amount by which the pretest probability is INCREASED in pts with POSITIVE test
SUPPORTS TRUE DISEASE; Should be slightly significant above 1. LR+ >10 causes large shift in posttest probability of disease
[(1-sensitivity)/specificity]; amount by which the pretest probability of disease is REDUCED in patients with a NEGATIVE test
Supports NO DISEASE; Should be significant
LR = 1
No shift in probability of disease
Predictive value importance
Sensitive to prevalence; indicates what the test MEANS to the patient
Positive predictive value
% of positives that are TRUE POSITIVES; [True positives/All positives]
Negative predictive value
% of negatives that are TRUE NEGATIVES; [True negatives/All negatives]
Shotgun; Many tests at once. ANY positive is diagnostic. INCREASES SENSITIVITY. Reduces risk from lack of intervention.