A Complicated PPV Question! - USMLE Forums
USMLE Forums Logo
USMLE Forums         Your Reliable USMLE Online Community     Members     Posts
Home
USMLE Articles
USMLE News
USMLE Polls
USMLE Books
USMLE Apps
Go Back   USMLE Forums > USMLE Step 1 Forum

USMLE Step 1 Forum USMLE Step 1 Discussion Forum: Let's talk about anything related to USMLE Step 1 exam


Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 11-29-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Stats A Complicated PPV Question!

A test yields 80% infected and 10% non-infected.....prevalence is 10%.
What is positive predictive value?

A. 47%
B. 54%
C. 75%
D. 25%
Reply With Quote Quick reply to this message



  #2  
Old 11-29-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 42
Threads: 10
Thanked 60 Times in 22 Posts
Reputation: 70
Default

A. 47%

Assuming total population of 100. 10% prevalence would mean total diseased number to be 10 out of which 80%(8) are ruled + by test. And so on. 8+9/17=0.47


Could you give the answer with possible explanation if I'm headed in the wrong direction.
Reply With Quote Quick reply to this message
  #3  
Old 11-29-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

sure i will let you know answer with explanation..waiting for some more answers...
Can you please explain how you calculated specificity??
Reply With Quote Quick reply to this message
 
  #4  
Old 11-29-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 42
Threads: 10
Thanked 60 Times in 22 Posts
Reputation: 70
Default

Okay. 1. I think I calculated something else in previous reply so ignore that. 2. Could you please write the whole question because either you wrote the question wrong or the answer I'm getting from your given values is not any one of the options.
Thank you.
Reply With Quote Quick reply to this message
  #5  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 65
Threads: 9
Thanked 14 Times in 9 Posts
Reputation: 24
Default

Same here just as insomniac did. I also got a different answer somewhere around 80%
Reply With Quote Quick reply to this message
  #6  
Old 11-30-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

New stool test for shigella infection yields postive results in 80% of infected patients and in 10% of uninfected patients ...prevalence of shigela in population is 10%..What is the probability that a patient who tests positive with the new test is infected with shigella ..

P.s: insomiac u were right BUT 8+9/17 = 0.47 .plz explain this equation.
Reply With Quote Quick reply to this message
  #7  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 11
Threads: 0
Thanked 6 Times in 4 Posts
Reputation: 16
Default

Quote:
Originally Posted by medicalbiology View Post
Question:
A test yields 80% infected and 10% noninfected.....prevalence is 10%
...what is positive predicted value ??

a.47%
b.54%
c.75%
d.25%

very confusing question!

if i suppose 80% infected as sensitivity and 10% noninfected as specificity then with a prevalnce of 10% , the positive predictive value is 8%.

now,if the question is correctly asked then i must have misinterpreted it.

'test yields 80% infected' means all the positive results ( TP and FP included ), then 80% is not the sensitivity n our whole calculation collapses n i can't find the answer then without sensitivity n specificity.
Reply With Quote Quick reply to this message
  #8  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 11
Threads: 0
Thanked 6 Times in 4 Posts
Reputation: 16
Default

Quote:
Originally Posted by medicalbiology View Post
New stool test for shigella infection yields postive results in 80% of infected patients and in 10% of uninfected patients ...prevalence of shigela in population is 10%..What is the probability that a patient who tests positive with the new test is infected with shigella ..

P.s: insomiac u were right BUT 8+9/17 = 0.47 .plz explain this equation.
equation is simple, Insomniac has wrongly put 9 in the numerator.


with 80% sensitivity n 10% specificity, using 2 by 2 table,
TP=80 , FP=90
FN=20, TN=10

PPV = 80/80+90 = 80/170 = 47%
Reply With Quote Quick reply to this message
  #9  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 11
Threads: 0
Thanked 6 Times in 4 Posts
Reputation: 16
Default

Quote:
Originally Posted by medicalbiology View Post
New stool test for shigella infection yields postive results in 80% of infected patients and in 10% of uninfected patients ...prevalence of shigela in population is 10%..What is the probability that a patient who tests positive with the new test is infected with shigella ..

P.s: insomiac u were right BUT 8+9/17 = 0.47 .plz explain this equation.

i am using this formula only using sensitivity and specificity,not putting 10% prevalence in consideration.

if using 10% prevalence in the formula, i get a diferent answer!

plz xplain medicalbiology!
Reply With Quote Quick reply to this message
  #10  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 45
Threads: 9
Thanked 12 Times in 5 Posts
Reputation: 22
Default

Quote:
Originally Posted by organo14 View Post
equation is simple, Insomniac has wrongly put 9 in the numerator.


with 80% sensitivity n 10% specificity, using 2 by 2 table,
TP=80 , FP=90
FN=20, TN=10

PPV = 80/80+90 = 80/170 = 47%
If this should b the case thn FP must equal to 10 not 90,
as q says test is positive in. 10 % who dnt hav disease
consequentlt equation become
80/80+10= 88%

If i m wrong correct me plz??
Reply With Quote Quick reply to this message
  #11  
Old 11-30-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

organ14 i am also confused in specificty thing..... According to statement it should be 10% specificity .. I willl let u know uworld explanation soon but explanation is also not very helpful...
Reply With Quote Quick reply to this message
  #12  
Old 11-30-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

Quote:
Originally Posted by hardtime View Post
If this should b the case thn FP must equal to 10 not 90,
as q says test is positive in. 10 % who dnt hav disease
consequentlt equation become
80/80+10= 88%

If i m wrong correct me plz??
Exactly hardtime FP should be 10 NOT 90..
Reply With Quote Quick reply to this message
  #13  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 11
Threads: 0
Thanked 6 Times in 4 Posts
Reputation: 16
Default

the question is wrongly put,

the specificity = 90%
sensitivity = 80%
prevalence of infection = 10%
suppose we have 1000 pts,
then

infected noninfected

test +ve 80 90 = 170
test -ve 20 890 = 910

100 900 = 1000


so PPV = 80/170 = 47%
Reply With Quote Quick reply to this message
The above post was thanked by:
devareddy (11-30-2013), dr mumble (09-12-2014), nsesereso (12-02-2013)
  #14  
Old 11-30-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

Answer is:
Quote:
Imagine that population consist of 100 patients .Since the disease prevalence is 10% that means 10 patients have disease and 90 do not ..Performing a test with with 80% sensiitivity on 10 diseased patients yields 8 TF . performing a test with 90% specificity on 90patients without disease yield 9 FP ..
PPV = 8/(8+9) = 47%
BUT same question how specificity is 90% ?
Reply With Quote Quick reply to this message
  #15  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 45
Threads: 9
Thanked 12 Times in 5 Posts
Reputation: 22
Default

Quote:
Originally Posted by medicalbiology View Post
Answer is:


BUT same question how specificity is 90% ?
Got it


As question asks 10% prevelance

sensitivity = 80 %

FP as question mentioned = 10% = and we know that 1-specificity=FP r specificity = 1-FP rates

So putting values = 1 - .1 = .9 r 90/100


We make square

Disease present disease absent
test positive 8 (TP) 9 (FP)

test negitive 2 (FN) 81 (FN)

10 90 100

By putting values we get the ans
Reply With Quote Quick reply to this message



  #16  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 42
Threads: 10
Thanked 60 Times in 22 Posts
Reputation: 70
Thumbs Up Here's the explanation.

Seems like I guessed correctlywhat medicalbiology was trying to say in the first post.
Here's how I had gotten around it.
Attached Thumbnails
A Complicated PPV Question!-usmle-forum-biostat-qs.jpg  
Reply With Quote Quick reply to this message
The above post was thanked by:
medicalbiology (12-01-2013), nsesereso (12-02-2013), Subuhi (12-02-2013)
  #17  
Old 11-30-2013
USMLE Forums Guru
 
Steps History: 1+CK+CS
Posts: 367
Threads: 0
Thanked 70 Times in 60 Posts
Reputation: 80
Default

Quote:
Originally Posted by Insomniac View Post
Seems like I guessed correctlywhat medicalbiology was trying to say in the first post.
Here's how I had gotten around it.
but how do we know that 10 percent were labeled as false positive...
Reply With Quote Quick reply to this message
  #18  
Old 11-30-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 42
Threads: 10
Thanked 60 Times in 22 Posts
Reputation: 70
Default

Quote:
Originally Posted by CisternaChyli View Post
but how do we know that 10 percent were labeled as false positive...
@CisternaChyli

It states that right in the question.
"New stool test for shigella infection yields postive results....."in 10% of uninfected patients""
Reply With Quote Quick reply to this message
  #19  
Old 11-30-2013
USMLE Forums Guru
 
Steps History: 1+CK+CS
Posts: 367
Threads: 0
Thanked 70 Times in 60 Posts
Reputation: 80
Default

Quote:
Originally Posted by Insomniac View Post
@CisternaChyli

It states that right in the question.
"New stool test for shigella infection yields postive results....."in 10% of uninfected patients""
whoops my bad, i didnt read carefully . It all makes sense now, just gotta read carefully. I get tripped up cause i forget to just assume the amount of patients.
Reply With Quote Quick reply to this message
  #20  
Old 12-01-2013
val7's Avatar
USMLE Forums Master
 
Steps History: CS Only
Posts: 557
Threads: 29
Thanked 296 Times in 186 Posts
Reputation: 316
Default

Quote:
Originally Posted by medicalbiology View Post
A test yields 80% infected and 10% non-infected.....prevalence is 10%.
What is positive predictive value?

A. 47%
B. 54%
C. 75%
D. 25%

It would be really helpful if you could tell us where this question is from, because if it is from NBME I won't look at it, same goes for uworld
Secondly, please just post the whole question next time.
The way you put it at first is so confusing, there is no way I could have known what this question is asking...
Please everyone who posts questions, lets try to do it the right way so that it helps us all!
__________________
Step1 [ ], Step 2CK [ ], Step 2 CS [Pass first attempt], Step 3 [ ]
Reply With Quote Quick reply to this message
  #21  
Old 12-01-2013
USMLE Forums Addict
 
Steps History: Not yet
Posts: 148
Threads: 44
Thanked 21 Times in 18 Posts
Reputation: 31
Default

I have a question...


what if the specificity and sensitivity is the same... ex. 45% and prevalence is 30%.

Then asked to calculate prevalence.. what would the prevalence be in this example?
Reply With Quote Quick reply to this message
  #22  
Old 12-01-2013
medicalbiology's Avatar
USMLE Forums Guru
 
Steps History: Not yet
Posts: 394
Threads: 128
Thanked 108 Times in 88 Posts
Reputation: 118
Default

val7 itz from uworld....Actually it is not allowed to write whole question of copyright material..
Reply With Quote Quick reply to this message
  #23  
Old 12-01-2013
USMLE Forums Veteran
 
Steps History: Not yet
Posts: 206
Threads: 13
Thanked 178 Times in 77 Posts
Reputation: 188
Default

PPV = TP/(TP+FP). It was impossible to work out the TP or FP rate from the way the original question was worded.

Sky_blue2000, your question is confusing, If the prevalence is 30%, then the prevalence is 30%
Reply With Quote Quick reply to this message
  #24  
Old 12-01-2013
USMLE Forums Addict
 
Steps History: Not yet
Posts: 148
Threads: 44
Thanked 21 Times in 18 Posts
Reputation: 31
Default

Oh sorry!!

I meant what would the PPV be if the prevalence is 30%

and specificity and sensitivity are both 45%
Reply With Quote Quick reply to this message
  #25  
Old 12-01-2013
USMLE Forums Veteran
 
Steps History: Not yet
Posts: 206
Threads: 13
Thanked 178 Times in 77 Posts
Reputation: 188
Default

Lets say sample size is 1000 patients

Prevalance = 30%

Therefore diseased = 300, non diseased = 700

We now need to work out our TP and FP numbers

Sensitivity = 45% In other words this test will give a positive result in 45% of the people with the disease (ie its rubbish!)

So the True Positive rate = 300 x 45% = 135.

This means that the False Negative rate, which is all the patients our test missed, is 300 - 135 = 165

Now, specificity = 45%

This means that, out of all the 700 patients without disease, our test will give a truly negative result only 45% of the time

So our True Negative rate = 700 x 45% = 315

This means that the False Positive rate, which is all the patients without disease that our test gave a positive result for, is 700 - 315 = 385

We now have all the numbers needed to make a 4x4 table and work out anything else.

The PPV = TP/(TP+FP) = 135/(135+385) = 26%

So the test is worse than simply tossing a coin...
Reply With Quote Quick reply to this message
The above post was thanked by:
Doc889 (01-23-2014), medicalbiology (12-01-2013), sky_blue2000 (12-01-2013)



  #26  
Old 12-01-2013
USMLE Forums Addict
 
Steps History: Not yet
Posts: 148
Threads: 44
Thanked 21 Times in 18 Posts
Reputation: 31
Default

Quote:
Originally Posted by USMLEUK. View Post
Lets say sample size is 1000 patients

Prevalance = 30%

Therefore diseased = 300, non diseased = 700

We now need to work out our TP and FP numbers

Sensitivity = 45% In other words this test will give a positive result in 45% of the people with the disease (ie its rubbish!)

So the True Positive rate = 300 x 45% = 135.

This means that the False Negative rate, which is all the patients our test missed, is 300 - 135 = 165

Now, specificity = 45%

This means that, out of all the 700 patients without disease, our test will give a truly negative result only 45% of the time

So our True Negative rate = 700 x 45% = 315

This means that the False Positive rate, which is all the patients without disease that our test gave a positive result for, is 700 - 315 = 385

We now have all the numbers needed to make a 4x4 table and work out anything else.

The PPV = TP/(TP+FP) = 135/(135+385) = 26%

So the test is worse than simply tossing a coin...

Thank you soo much!! Great explanation
Reply With Quote Quick reply to this message
  #27  
Old 12-01-2013
USMLE Forums Scout
 
Steps History: Not yet
Posts: 42
Threads: 10
Thanked 60 Times in 22 Posts
Reputation: 70
Cool Waoh!, that was a pain.:p

Quote:
Originally Posted by sky_blue2000 View Post
Oh sorry!!

I meant what would the PPV be if the prevalence is 30%

and specificity and sensitivity are both 45%

Applying Bayes Theorem:

PPV = Sensitivity x Prevalence/ [ (Sensitivity x Prevalence) + (1-Specificity)(1-Prevalence)]

Plug in those values and you get ~26% PPV
Reply With Quote Quick reply to this message
The above post was thanked by:
Doc889 (01-23-2014), USMLEUK. (12-01-2013)
  #28  
Old 12-01-2013
USMLE Forums Veteran
 
Steps History: Not yet
Posts: 206
Threads: 13
Thanked 178 Times in 77 Posts
Reputation: 188
Default

Ha, i remember seeing that on uworld and clearly totally didnt take it in...
Reply With Quote Quick reply to this message
  #29  
Old 12-01-2013
faith255's Avatar
USMLE Forums Master
 
Steps History: 1+CK+CS
Posts: 521
Threads: 41
Thanked 74 Times in 63 Posts
Reputation: 84
Help confusion..

Quote:
Originally Posted by USMLEUK. View Post
Lets say sample size is 1000 patients

Prevalance = 30%

Therefore diseased = 300, non diseased = 700

We now need to work out our TP and FP numbers

Sensitivity = 45% In other words this test will give a positive result in 45% of the people with the disease (ie its rubbish!)

So the True Positive rate = 300 x 45% = 135.

This means that the False Negative rate, which is all the patients our test missed, is 300 - 135 = 165

Now, specificity = 45%

This means that, out of all the 700 patients without disease, our test will give a truly negative result only 45% of the time

So our True Negative rate = 700 x 45% = 315

This means that the False Positive rate, which is all the patients without disease that our test gave a positive result for, is 700 - 315 = 385

We now have all the numbers needed to make a 4x4 table and work out anything else.

The PPV = TP/(TP+FP) = 135/(135+385) = 26%

So the test is worse than simply tossing a coin...
u xplained it really well but i have 1 confusion.. in calculating the TP u did 300 x 45% .. thn did u divide it by 100 to get rid of the % ??

i knw my ques is lame but jst want to b sure.. weak in maths :/

thanks..
__________________
Never let your fear be stronger than your Faith
Reply With Quote Quick reply to this message
  #30  
Old 12-02-2013
USMLE Forums Veteran
 
Steps History: Not yet
Posts: 206
Threads: 13
Thanked 178 Times in 77 Posts
Reputation: 188
Default

yep, multiplying something by 45% (per-cent) just means multiplying by "45 per 100" ie 45/100, or 0.45
Reply With Quote Quick reply to this message
  #31  
Old 12-02-2013
USMLE Forums Master
 
Steps History: 1+CK+CS+3
Posts: 1,109
Threads: 31
Thanked 521 Times in 394 Posts
Reputation: 540
Default

This same question was discussed in the beh science vedio of Kaplan. The answer is 47%. Insomnia's photograph explanation is correct!
__________________
In USMLE, “Any fool can know. The point is to understand.”
Reply With Quote Quick reply to this message



Reply

Tags
Biostatistics-Epidemiology, Step-1-Questions

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the USMLE Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:
Medical School
Choose "---" if you don't want to tell. AMG for US & Canadian medical schools. IMG for all other medical schools.
USMLE Steps History
What steps finished! Example: 1+CK+CS+3 = Passed Step 1, Step 2 CK, Step 2 CS, and Step 3.

Choose "---" if you don't want to tell.

Favorite USMLE Books
What USMLE books you really think are useful. Leave blank if you don't want to tell.
Location
Where you live. Leave blank if you don't want to tell.

Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.



Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Last Post
Complicated CF Genetics Pedigree ashanair70 USMLE Step 1 Forum 21 02-20-2013 02:51 PM
Grief versus Complicated Grief versus Major Depression winner21 USMLE Step 2 CK Forum 2 05-05-2012 09:51 AM
What is the diagnosis of this complicated patient? m82_ghasemi USMLE Step 1 Forum 9 09-27-2011 11:48 AM
A complicated case of a smoker, carpenter, and alcoholic man! aghammoud85 USMLE Step 2 CK Forum 4 05-31-2011 03:00 PM
Complicated unlicensed abortion RRMadukha USMLE Step 2 CK Forum 1 05-26-2010 02:49 PM

RSS Feed
Find Us on Facebook
vBulletin Security provided by vBSecurity v2.2.2 (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.

USMLE® & other trade marks belong to their respective owners, read full disclaimer
USMLE Forums created under Creative Commons 3.0 License. (2009-2014)