EAB: Diagnostic Tests and Screening Flashcards
Define validity.
Validity = truth of result, lack of bias
Define reliability.
Reliability = consistency of results
How can you improve reliability in a test?
- repeatability
- reproducibility
What is sensitivity?
Sensitivity = TP / (TP+FN)
Proportion with disease who test positive.
what is specificity?
Specificity = TN / (TN+FP)
Proportion without disease who test negative
What is the positive predictive value?
PPV = TP / (TP+FP)
Proportion with positive test who have disease
What is the ROC curve?
- sensitivity % (TP) on y axis plotted against 100-specificity % (FP) on x axis
- useful test should have plots above 45 degrees
- improved by combining different tests together
what test do we use for this
“if someone has the disease, what is the probability that the test will be positive?”
sensitivity
what is the test for this:
“if someone does not have the disease, what is the probability that the test will be negative?”
specificity
is screening a form of primary or secondary prevention?
secondary
what is the test for this:
“if someone has a positive test, what is the probability that they will have the disease?”
positive predictive value
what is the difference to primary and secondary prevention?
- primary prevention: aiming to delay disease onset in healthy people
- secondary intervention: aims to reduce disease’s impact (AFTER DISEASE ONSET) - FOR EXAMPLE SCREENING IS AN EXAMPLE OF SECONDARY PREVENTION AS ITS DONE AFTER DISEASE ONSET
What are the 5 criteria for introducing a screening programme?
- Condition
- Test
- Treatment
- Screening programme
- Role of UK National Screening Committee (NSC)
What are 5 conditions for screening?
- Criterion 1: the condition is important in terms of frequency, impact on affected individuals
- Criterion 2: Primary prevention is not effective or feasible
- Criterion 3: There is an identifiable pre-clinical stage of disease
- Criterion 4: There is a test that can separate those with a high probability of the disease from those with a low probability
-
Criterion 5: Effective treatment is available; the
outcome of early treatment is better than later
treatment after clinical diagnosis
What are 4 characteristics of the screening test?
- The test should be simple, safe, precise and have been validated
- There should be a clear distinction between ‘normal’ and ‘abnormal’ results
- The test should be acceptable to subjects, and costs should be ‘reasonable’
- There should be an agreed policy on the further diagnostic investigation of individuals with a positive test result and on the choices available
What is lead time bias?
interval between the diagnosis of a disease at screening and the usual time of diagnosis (by symptoms)
IE time between:
diagnosis by screening —> diagnosis by symptoms
What is length bias?
sampling - rapidly progressive disease cause the individual to consult, but less rapidly progressing cases are likely to remain for screening detection
THEREFORE slow progressing diseases are more often found by screening and have better chances of being treated
what is selection bias?
those who enter screening almost invariably are more conscious than those who decline
what is overdiagnosis
bias - some lesions identified as disease in a screening programme would not have presented clinically during the individual’s life time
what are the 4 biases associated with screening
lead time
length biased
selection bias
overdiagnosis bias
what are 3 solutions to biases
- Randomised control trials when mortality rather than survival is used as the outcome
- Survival can be used only if there is no evidence of overdiagnosis bias and observation period is from randomisation date.
- Individual and community trials
What are the 4 advantages of screening?
- Improved prognosis for true positives
- Less radical treatment required
- May be resource savings
- Reassurance for those with true negative test results
What are the 5 disadvantages of screening?
- Longer period of awareness for true positives
whose prognosis is unaltered - Over-treatment of borderline abnormalities
- False reassurance of false negatives
- Anxiety and hazard for false positives
- Hazard of screening test to all recipients
whats the difference between sensitivity and the positive predictive value?
sensitivity = proportion WITH DISEASE who test positive
PPV = proportion WITH POSITIVE TEST who have disease
