EIP: Screening and Diagnostic Tests - Week 12 Flashcards
A visit to the optometrist for a patient over 40 usually means an IOP measurement. Why?
Glaucoma. Increased risk of Glaucoma in over 40yrs old. Use IOP measurement as a value to compare to the normal value (21mmHg is upper limit of normal. If over 21, investigate why)
- use value to determine whether it’s more likely or not that they have glaucoma and whether you have to investigate further
What other screening activities does an optometrist perform? (other than IOP measurement) Name 1 example
Confrontation
What are screening tests used for? Generally
Identifying patients that need further investigation
Define Screening Tests (UK National Health Service definition)
a public health service in which members of a defined population, who do not necessarily perceive they are at risk of, or are already affected by a disease or its complications, are asked a question or offered a test, to identify those individuals who are more likely to be helped than harmed by further tests or treatment to reduce the risk of a disease or its complications
What does the effectiveness of a screening test depend on? (generally) (3)
the characteristics of the test
the disease being screened
the population being screened
What characteristics should a screening test have? (5)
fast
easy
cheap
reliable (high sensitivity and specificity) (note: sensitivity is the chance of the test being positive if they have the disease)
painless
What features of the disease are important to a screening test? (why is screening useful for disease?) (4)
better prognosis
no other signs
prevalance of the condition
Impact of the disease
The people screened should be ….? (2)
high risk group
should be accepting of the test
What are the downsides of screening? (4)
expensive
lot of effort + coordination
false positives can be a problem (people flagged with a condition without having it will stress the patient and cost them a lot in tests)
false negatives (false sense of security)
When should screening only be introduced?
When the balance of benefits vs harms and costs is favourable, and this has been demonstrated by sound evaluation - “signal detection theory”
What does it mean if a test is 100% sensitive?
always positive in subjects in whom the disease of interest is present (i.e. every case is picked up)
What does it mean if a test is 100% specific?
only positive when the disease of interest is present
always negative when the disease is absent
Assuming a disease with 1% prevalence: Under what conditions would a test with more specificity or sensitivity be preferred?
Higher specificity: good if disease does not have significant impact, or if subsequent testing is difficult/expensive
Higher sensitivity: good if disease has a significant impact, or if subsequent testing is easy/cheap
Define Positive Predictive Value (PPV)
The probability of having the disease when the test is positive
(this is exactly what you want to know when you screen someone)
Define Negative Predictive Value (NPV)
The probability of not having the disease when the test is negative
How does prevalence of the disease affect PPV and NPV?
Large impact. High prevalence increases PPV and decreases NPV
What might pre-test probability estimates depend on? (5)
our own clinical experience in detecting symptoms and signs of the disease
regional or national prevalence statistics
databases
the accuracy and importance of the test shown in an original report on the test
specific studies that determine the pre-test probabilities
What does a likelihood ratio for a positive test result show you?
shows you how much more likely a patient’s positive test result would be for someone with the disease compared to someone without the disease
LR+ = sensitivity/(1-specificity)
What does a likelihood ratio for a negative test result show you?
shows you how much more likely a patient’s negative test result would be for someone without the disease compared to someone with the disease
LR- = (1-sensitivity)/specificity
What does the Receiver Operating Characteristic (ROC) curve plot?
sensitivity as a function of 1-specificity for different criteria
What is the ideal area under the curve (AUC) for the ROC curve?
1
What is the worst AUC for the ROC curve?
0.5
T/F: Increasing AUC of the ROC curve makes the test better
True
