|
Author
|
Message
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
screening confusa!
|
06.21.04 (4 years ago)
#1
|
|
This question is taken from RxPG AIPGE
'04 Book:
| Quote: |
Q.The usefulness of a 'screening test' in a community depends on its:
a)sensitivity
b)specificity
c)reliability
d)predictive value
|
ans given is d) predictive value
am sorta confused between 'sensitivity' & 'predicitive value'.
watever I could make out from reading related books are:
[] For a screening test we use a test which is 'highly sensitive'.
[] A screening test is more based on the theory of 'ruling out' a disease than confirming its presence.(i.e. if the result of a highly sensitive test is negative,it allows the disease to be ruled out with confidence!).
[] While an epidemologist is more concerned with the question of 'the ability of a test to detect the presence/absence of a disease'(ie the sensitivity of a test),
a physician is more concerned with the question of "How likely is it that the disease in question is actually present/absent?" (ie the predictive value of the test). 
[] so when 'sensitivity' wud describe the 'usefulness' of a screening test,
'predictive value' will determine the usefulness of a diagnostic test.
I am sorry if all this sounds confusing.
but this is what i could conclude.I might be wrong.
Pls opine! 
all ref from High Yeild Biostats pg 72-74
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
|
|
Bruno
Credits:
45896
My Scrapbook
My Reading List
50 Books
|
|
|
06.21.04 (4 years ago)
#2
|
|
Please see the question
usefulness of a 'screening test' in a community
The question is not
Usefullness of the diagnostic test for an individual or for High Risk Group
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
Bruno
Credits:
45896
My Scrapbook
My Reading List
50 Books
|
|
|
06.21.04 (4 years ago)
#3
|
|
am sorta confused between 'sensitivity' & 'predicitive value'.
watever I could make out from reading related books are:
| Quote: |
| [] For a screening test we use a test which is 'highly sensitive'. |
Correct
| Quote: |
| [] A screening test is more based on the theory of 'ruling out' a disease than confirming its presence.(i.e. if the result of a highly sensitive test is negative,it allows the disease to be ruled out with confidence!). |
Correct
| Quote: |
| [] While an epidemologist is more concerned with the question of 'the ability of a test to detect the presence/absence of a disease'(ie the sensitivity of a test), a physician is more concerned with the question of "How likely is it that the disease in question is actually present/absent?" (ie the predictive value of the test). |
I beg to differ....
As a clinician, I am not bothered "How likely is it that the disease is actually present or absent".... because as a clinician I AM TREATING ONE PATIENT and when he is positive I go for a Confiramtory test..... A patient comes with a leg ulcer that is not healing..... I ask him to test his urine for sugar.... When that is positive, i order for a Fasting and Post Prandial Bllod Sugar......... And I am "concerned with the question of 'the ability of a test to detect the presence/absence of a disease'(ie the sensitivity of a test)"
A clinician treats ONE patient and he needs sensitivity as well as specificity
And the EPIDEMIOLOGIST has to treat a COMMUNITY........
| Quote: |
[] so when 'sensitivity' wud describe the 'usefulness' of a screening test,
'predictive value' will determine the usefulness of a diagnostic test. |
FOR AN INDIVIDUAL..........BY THE CLINICIAN
so when 'sensitivity' wud describe the 'usefulness' of a screening test,
'specificity' will determine the usefulness of a diagnostic test
See the next post for the Epidemiologist and Community
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
Bruno
Credits:
45896
My Scrapbook
My Reading List
50 Books
|
|
|
06.21.04 (4 years ago)
#4
|
|
Before Proceeding further, the reader may be benefitted if he read Chapter 4 (Screening) from Park...... Those who cannot  (browsing centre) or will not  (lazy fellow like me) be sure of one fact
Positive Predictive Value of a test in a Particular Setting Increases as the Prevalance Increases
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
Bruno
Credits:
45896
My Scrapbook
My Reading List
50 Books
|
|
|
06.21.04 (4 years ago)
#5
|
|
Now tell me the answer for this
A1. You are Screening for HIV at 1000 children taken from a nprimary school
A2. You are screening for HIV at 1000 lorry drivers
Or see this
B1. You are screening for TB at 1000 out patients tommorow morning at Apollo Hospitals Medicine OP
B2. you are Screening for TB at 1000 patients at an AIDS Rehabilitation Centre
As an epidemiologist what do you need MORE in this case.... (You need everything.... that is a different fact)
1. Do you need a test that is sensitive
2. Do you need a test such that....... The majority of those who are positive really have the disease...... Remember that you are doing the test in 1000s..... . And for every one who is positive in the initial test, you have to do a screening (and that involves a lot of money)
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
|
|
06.22.04 (4 years ago)
#6
|
|
| Bruno wrote: |
| Quote: |
| [] While an epidemologist is more concerned with the question of 'the ability of a test to detect the presence/absence of a disease'(ie the sensitivity of a test), a physician is more concerned with the question of "How likely is it that the disease in question is actually present/absent?" (ie the predictive value of the test). |
I beg to differ.... |
well, I cant comment much on this,except that this is how High Yeild has tried to explain the difference between the predictive value & sensitivity.
pg 74 1st para under Predictive Values.
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
|
|
06.22.04 (4 years ago)
#7
|
|
| Quote: |
| A patient comes with a leg ulcer that is not healing..... I ask him to test his urine for sugar.... |
ie we 1st resort to a test which can roughly rule out the presence of DM
or
we r relying on 'the ability of a test to detect the presence/absence of a disease'(ie the sensitivity of a test),
| Quote: |
| When that is positive, i order for a Fasting and Post Prandial Bllod Sugar......... |
i.e. before starting the actual treatment we want to confirm the diagonis which was suggested by the 1st test,
i.e. we want to know "How likely is it that the disease in question is actually present/absent?" (ie the predictive value of the test).
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
|
|
06.22.04 (4 years ago)
#8
|
|
| Quote: |
| Positive Predictive Value of a test in a Particular Setting Increases as the Prevalance Increases |
true!
(PPV = true positives/true positives+ false positives)
doesnt this means that for a rare disease the PPV will be falsely very low(as it wud give a high number of false positives).
i.e. this variable called PPV of a test, fluctuates not only with the type of test,but also with the setting/population on which it is applied.
how useful can be a variable in a screening test which fluctuates with the type of disease & the type of population on which it is applied?
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
|
|
06.22.04 (4 years ago)
#9
|
|
in contrast when come to sensitivity of a test,
sensitivity determines only the proportion of the persons who test positive for that test to the total number of the disease.
a highly sensitive test wud mean that a large number of people with the disease are testing actually positive & only a few are left behind.
where as a test with high PPV may not necessarily mean this.
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
akanksha
Credits:
41289
My Scrapbook
My Reading List
14 Books
|
|
|
06.22.04 (4 years ago)
#10
|
|
sensitivity is the 'catching power' of a test ie describes how efficiently can a test accurately detect the disease.
& it depends on the test itself.
if a test is 90% sensitive,
it remains 90% sensitive whether we apply this test to one person or 1000 persons,
whether the disease is rare in the particular population,
or,
whether we r applying the test in a hospital setting or general population.
wheras same doesnt holds true for PPV!
|
|
|
Post Options:
Add to favourite
.
Tell a friend
.
|
|
Back to top
|
|
|
Topic Tags: 
