Scribd
Upload
176 views8 pages

Bayes Theorem: HIV Test Example

This document provides an example of using Bayes' theorem to calculate conditional probabilities for the results of a medical test. Specifically: - 15% of patients have HIV, and the test correctly identifies 95% of HIV+ patients and incorrectly identifies 2% of HIV- patients as positive. - If a patient tests positive, the probability they have HIV is 89.34% and the probability they do not is 1.07%. - If a patient tests negative, the probability they have HIV is 0.89% and the probability they do not is 99.11%.

Uploaded by

Ankitjee2007
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
176 views8 pages

Bayes Theorem: HIV Test Example

This document provides an example of using Bayes' theorem to calculate conditional probabilities for the results of a medical test. Specifically: - 15% of patients have HIV, and the test correctly identifies 95% of HIV+ patients and incorrectly identifies 2% of HIV- patients as positive. - If a patient tests positive, the probability they have HIV is 89.34% and the probability they do not is 1.07%. - If a patient tests negative, the probability they have HIV is 0.89% and the probability they do not is 99.11%.

Uploaded by

Ankitjee2007
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

BAYES THEORAM

Example
Example
• Lets consider another example of the use of Bayes Theorem. This time we are
given different information from the previous example.

• In a certain clinic 0.15 of the patients have got the HIV virus. Suppose a blood test
is carried out on a patient. If the patient has got the virus the test will turn out
positive with probability 0.95. If the patient does not have the virus the test will
turn out positive with probability 0.02.

• If the test is positive what are the probabilities that the patient
a) has the virus
b) does not have the virus?

• If the test is negative what are the probabilities that the patient
c) has the virus
d) does not have the virus?
• Lets give the events the following labels
H = the patient has got the virus
P = the outcome of the test is positive

• The question above gives us the following information.


P(H) = 0.15
P(P/H)=0.95
P(P/H’)=0.02

• And we are asked to find the following


a) P(H | P)
b) P(H’/P)
c) P(H’/P’)
d) P(H’/P’)
• To find the first of these, P(H|P), we can write
down Bayes Theorem
P(H/P) = P(P/H).P(H)
P(P)
• But we notice that we are given two of the
values on the right-hand side P(P|H) and P(H)
but we are not given P(P) – the probability of a
positive result. So we have to work this out
ourselves.
• Now there are two ways that a patient could
have a positive result:
– he has the virus and he gets a positive result
– he does not have the virus and he gets a positive
result

• We have to work out the probabilities of both


these cases and add them together.
P(P)=P(H^P)+P(H’^P)
• Now from the second axiom of probability we
have
P(H^P)=P(P/H)P(H)
=0.95*0.15=0.1425
P(H’^P)=P(P/H’)P(H)
= 0.02*0.85 = 0.017
SO
P(P)=0.1425+0.017=0.1595
P(H/P) = P(P/H).P(H)
P(P)
P(H/P) = 0.95 × 0.15 / 0.1595
= 0.8934
(B) P(H’/P)
P(H’/P)=1-P(H/P)
= 1- 0.8934
= 0.01066
( C ) P(H/P’)
P(H/P’)= P(P’/H)P(H)
P(P’)
= (0.05 * 0.15) / ( 1 - 0.1595)
= 0.0089233
(D) P(H/P)
P(H/P)=1-P(H/P’)
=1-0.0089233
=0.99107

You might also like