4: Medical diagnosis

Question 1

Medical diagnosis

Answer 1

• Process of determining the health status and the factors responsible for producing it
• classification of an individuals conitions into seperate and distinct categories that allow medical decisions about treatment and prognosis
• motivated by presense of signs

Question 2

Screening

Answer 2

• Identification of unrecognized disease / defect by application of tests
• adresses well persons who could still have the disease, so therefore not intended to be diagnostic

Question 3

Gold standard

Answer 3

• Not the truth, but the cloest one can get ( best available option)
• all other tests compared to it, but often necessary to develop other test cause gold standard is to complex, expensive, …
• Problem: lack of gold standard, or gold standard imperfect itself

Question 4

Validity of a test

Answer 4

Degree the tool measures what it claims to measure

Sensitivity
How many true positives are tested positive?

Specificity
How many true negatives are tested negative?

Question 5

Specificity and Sensitivity
contigency table
formula
standard sentence

Answer 5

Sensitivity
X% of the truly ill or true positives get a positive test result

Specificity
X% of the trully healthy or true negatives get a negative test result.

Question 6

When sensitivity or specificity is very important to be close to 100%?

Disease
Treatment
False positive results
Example

Answer 6

Question 7

What do you need to apply if there is not only 2 categories (yes or no)?

Answer 7

Polytomous test result like urea breath test require a cut-off value, that has to consider both sensitivity and specificity

Therefore are Reiceiver-Operating-Characteristics Curve is used to identifiy the proper cut off value

Question 8

What does a diagnostic test tell me?

Diagnostic test measures and what do you have to consider ?

Answer 8

Positive predictive value
The probability of being ill when the result is positive

Negative predicitve value
The probability of being healthy when the test is negative

These values depend on the prevalence, the higher the prevalence, the higher the PPV and the lower the NPV

Question 9

Likelihood Ratio
LR+
LR-
standard sentence
Interpretation

Answer 9

LR determines wheter a test result usefully updates the probability that a condition exisit

LR+: How much more likely is it that
- a person with disease has a positive test result than
- a person without disease has a positive test result

LR-: How much more likely is it that
- a person with disease has a negative test result than
- a person without disease has a negative test result

Standard sentence
The likelihood that a person with disease get a given test result is X-times as large as the likelihood for the person without the disease to get the same test result

Interpretation
LR+ > 10 = excelent
LR+ = 1-2 = bad

LR- < 0.1 = excelent
LR- = 0.5-1 = bad

Question 10

Pre and post test probability
steps to determine
easy way to calc Post test probability

Answer 10

Pre test probability
Probability of disease among those you attend for examination (often = prevalence)

Prevalence = p

Pre-test Odds
Odds = p/(1-p)

Post-test Odds
Pre-test Odds * LR = Post-test Odds

Post test probability
probability of disease with positive test result

p = Odds/(1+Odds)

easy way to calc Post test probability
Fagan nomogram with prevalence on the left, likelihood in the middle and post test probability on the right, you the post test probability if you draw a line from pre-test probability through LR to the post-test probability scale, but hatd to read exact numbers