CK040 - Diagnosis Flashcards
What is the odds at the ‘no treat-treat threshold’ ?
Harm / Benefit
What is the probability at the ‘no treat-treat threshold’ ?
Harm / (Harm + Benefit)
What is the ‘net harm’ ?
Treatment if no disease (so ‘difference in outcome with/without treatment if disease is NOT present’)
What is the ‘net benefit’ ?
Treatment if disease (so ‘difference in outcome with/without treatment if disease is present’)
What is the ‘sensitivity’ ?
True Positive Rate
(so the probability of getting a positive test result, GIVEN disease is present)
What is the ‘specificity’ ?
True Negative Rate
(so the probability of getting a negative test result, GIVEN disease is NOT present)
What is the ‘PPV’ ?
Probability of having the disease, GIVEN a positive test result
What is the ‘NPV’ ?
Probability of having NO disease, GIVEN a negative test result
If you want to make sure you do not miss any diagnosis, what should be high?
Sensitivity
What is the Bayes’s formula for tests?
Posterior odds = Prior odds x Likelihood Ratio
What are the formulas for the ‘Likelihood Ratio’ ?
- p(R | D+) / p(R | D-), where R is a certain test result
- TPR / FPR for a positive test result
- FNR / TNR for a negative test result
What does a ‘LR > 1’ mean?
That you are more likely to find this test result in diseased
What does a ‘LR < 1’ mean?
That you are more likely to find this test result in non-diseased
What happens to the thresholds when the specificity gets bigger?
‘No treat-test threshold’ gets smaller
What happens to the thresholds when the sensitivity gets bigger?
‘Test-treat threshold’ gets smaller
What is on the axis of a ‘ROC curve’ ?
- x-axis –> FPR (1 - specificity)
- y-axis –> TPR (sensitivity)
What is the KEY step in constructing a ‘ROC curve’ ?
Ordering the test results according to increasing Likelihood Ratio
What is the assumption of ‘conditional independence’ ?
Both for presence & absence of disease, the probability of combination of results is the same as the product of the results independenctly
When combining tests, what happens to the ‘sensitivity’ and ‘specificity’ when we us ‘OR’ ?
- Sensitivity goes up
- Specificity goes down
When combining tests, what happens to the ‘sensitivity’ and ‘specificity’ when we us ‘AND’ ?
- Sensitivity goes down
- Specificity goes up
What is ‘calibration’ ?
Is our model predicting what we think it should?
How to check for ‘calibration’ ?
Plot the predicted probabilities (x-axis) against the observed probabilities (y-axis). This line should be linear, and go through the origin.
What is ‘discrimination’ ?
How good does the prediction model discriminate diseased from non-diseased?
How to check for ‘discrimination’ ?
Use the AUC of the ROC-curve
What happens on the ROC curve when there is a high ‘prior odds’?
You move to the top right (because the LR decreases)
What happens on the ROC curve when there is a ‘large loss due to FP’?
You move to the bottom left (because LR increases)
What happens on the ROC curve when there is a ‘large loss due to missed diagnosis’?
You move to the top right (because LR decreases)
What is ‘verification bias’ ?
Selective use of the reference test conditional (based) on the index test result
What is ‘test-review bias’ ?
When the interpretation of the results of a test is influenced by knowledge of prior tests
What is ‘uninterpretability bias’ ?
When tests that yield uninterpretable or inconclusive results are excluded from the analysis