Hard Biostatistics Questions. - 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 01-26-2016
USMLE Forums Newbie
 
Steps History: Not yet
Posts: 7
Threads: 3
Thanked 1 Time in 1 Post
Reputation: 11
Default Hard Biostatistics Questions.

A 48-year old man is referred for evaluation of possible hypertension. On the basis of ten measurements, the patient's average diastolic blood pressure is 113mmHg, and the standard deviation is 8mmHg. If four rather than ten measurements of the mean blood pressure are made, which of the following is the expected impact on the size of the 95% confidence interval about the mean blood pressure?

A) Change, but the direction cannot be predicted
B) Decrease in width
C) Increase in width
D) Remain the same

As far as I know, and allegedly, the answer is C.

But I chose A because you never know what would happen to 4 newly measured values' standard deviation. I mean, it was 8mmHg when you measured 10 times, there is possibility it would be 1mmHg when measured 4 times with minor probability. I also understand that you should use σ(the population standard deviation), but you can use s(standard deviation within sample) instead where you can't find σ, which happens almost always in the field of inferential statistics including this situation.

So if this is the case, the size of the 95% confidence interval decrease even if (root 10) is larger than (root 4), because 8mmHg is far larger than 1mmHg.
And of course there is possibility that 95% confidence interval remains same contrary to choice A, but the chance is nearly zero because blood pressure is interval scale. If you take any 2 real number which is interval scale between 0 and 1 , the chance that you get same 2 number is zero.
These were the reasons why I thought A is answer.

Please tell me what's the right way to think.
Reply With Quote Quick reply to this message
The above post was thanked by:
chanduM (03-12-2016)



  #2  
Old 01-26-2016
USMLE Forums Newbie
 
Steps History: Not yet
Posts: 7
Threads: 3
Thanked 1 Time in 1 Post
Reputation: 11
Default Second Hard Biostat Question

An 80-year-old woman is being evaluated for suspected temporal arteritis(TA). Her erythrocyte sedimentation rate(ESR) is 100mm/h. Treatment of TA involves the use of glucocorticoids, which could have serious adverse effects. The pretest probability for TA is 50% in this patient. In the evaluation of TA, ESR has a sensitivity of 99% and a specificity of 60%. Based on the results of the ESR testing in this patient, which of the following is the most appropriate next step in management?

A) Additional testing to confirm the diagnosis of TA
B) Corticosteroid therapy, since the diagnosis of TA has been established with 99% certainty
C) Elimination of TA from further diagnostic consideration
D) Repeat ESR; if again positive, corticosteroid therapy
E) Repeat ESR; if normal, additional testing to confirm the diagnosis of TA

My choice was D but A is allegedly correct answer.
I used Likelihood ratio and pretest probability assuming sensitivity is 1(which is almost the same as 99% and doesn't make much difference) to calculate posttest probability which was 5/7.(LR : 2.5, Pretest probability is 0.5 according to question stem). Below is the formula I used.
P' = P0*LR/(1-P0+P0*LR), where P0 is the pre-test probability, P' is the post-test probability, and LR is the likelihood ratio. This formula can be calculated algebraically by combining the steps in the preceding description.(quoted from WIKI 'likelihood ratio')
as P' is equal to positive predictive value, PPV is alreay 5/7. If you do it once again like option D and assume it comes out positive again, then I can calculate 2nd P', which I call P''(double prime) and is 25/29=86% (you use 5/7 instead of 0.5 this time in the position of P0 in formula above). I believe this number is great enough to diagnose Temporal Arteritis and give steroids to prevent blindness with all 'serious adverse effects' because the benefit outweighs the harm here.

Help me!!!
Reply With Quote Quick reply to this message
  #3  
Old 01-26-2016
USMLE Forums Guru
 
Steps History: 1 + CS
Posts: 305
Threads: 40
Thanked 103 Times in 64 Posts
Reputation: 113
Default

Well I understood the first question
Confidence interval= mean +/_ standard deviation
As u see the equation of SD the sample n comes in the denomitor
So as the same decreases the confidence interval width increases.......the range I mean

Eg if the mean is 100 and std dev was 10......so CI is now 90-110
But if the sample size decreases so that the SD is increased.....for eg to 20 then CI is now 80-120
So the range width increases

I'm not sure abt the second ques
Maybe it s because the specificity is low which actually rules in the disease? But I'm sure this is nt thr ryt explanation.....pathoma says we shud give steroids even if we suspect TCA.... And tat kind off and is there in the options to confuse me!! Lol!!
Reply With Quote Quick reply to this message
The above post was thanked by:
CHASER (01-27-2016)
 
  #4  
Old 01-26-2016
USMLE Forums Scout
 
Steps History: 1+CK+CS
Posts: 35
Threads: 1
Thanked 3 Times in 1 Post
Reputation: 13
Default

at the first question i had the same reasoning as you did...

At the second question I choose A but I cannot give you a logic pattern. I just thaught if I had a patient with onliy ESR of 100 would I treat for TA? no...60% it's just not enough for making a dagnosis and I don;t think it's high enough to start treatment
Reply With Quote Quick reply to this message
  #5  
Old 01-27-2016
USMLE Forums Scout
 
Steps History: Not yet
Posts: 86
Threads: 1
Thanked 50 Times in 35 Posts
Reputation: 60
Default

The answer for the second question is A in my opinion.

A positive ESR means that you can not rule out temporal arteritis (high sensitivity, meaning that if you take 100 patients with known TA, 99 of them are going to test positive on ESR) but since elevated ESR can be caused by a million things other than temporal arteritis you need some other more specific test to confirm it (biopsy?).

That would be an answer to a pure biostatistics question.

But I think I heard on pathoma that a suspicion of temporal arteritis is enough of a reason to start steroids even before you test for it since it can cause blindness (I do not know how correct that is). But B is incorrect since 99% sensitivity does not establish a diagnosis with 99% certainty. A 99% positive predictive value establishes a diagnosis with 99% certainty, not sensitivity.

C is not correct because you just tested positive on a highly sensitive test. Which by definition means you can not rule out a diagnosis (99% of people with known TA test positive on this test). A negative result on a sensitive test would eliminate the possibility of temporal arteritis (because only 1% of people with known TA test negative on this test).

D does not look right. I think the person who started this thread is overthinking the question. Those board style questions like to give a long stem with a crapload of unnecessary information to try and throw you off. Also, Step 1 will not require you to judge if 86% is a number high enough to diagnose TA. Doing the test again and getting a positive result does not change anything. The test still has low specificity. No matter how many times you do the test it will not become highly specific and it will never allow you to confirm a diagnosis. A second positive result on a test will tell you that based on this result you can not exclude the possibility of TA, the same thing that the first test result told you.

E does not make any sense. If ESR were normal that would rule out TA (again, only 1% of patients with known TA test negative on this test)

The fact that pretest probability is very high in this patient it increases PPV and decreases NPV. Which kind of messes with you, but still, you get about a 71% PPV and a 98% NPV with this pretest probability.
Reply With Quote Quick reply to this message
The above post was thanked by:
CHASER (01-27-2016), Kruno (03-11-2016)
  #6  
Old 01-27-2016
USMLE Forums Scout
 
Steps History: 1+CK+CS
Posts: 83
Threads: 2
Thanked 39 Times in 30 Posts
Reputation: 49
Default

Quote:
Originally Posted by CHASER View Post
An 80-year-old woman is being evaluated for suspected temporal arteritis(TA). Her erythrocyte sedimentation rate(ESR) is 100mm/h. Treatment of TA involves the use of glucocorticoids, which could have serious adverse effects. The pretest probability for TA is 50% in this patient. In the evaluation of TA, ESR has a sensitivity of 99% and a specificity of 60%. Based on the results of the ESR testing in this patient, which of the following is the most appropriate next step in management?

A) Additional testing to confirm the diagnosis of TA
B) Corticosteroid therapy, since the diagnosis of TA has been established with 99% certainty
C) Elimination of TA from further diagnostic consideration
D) Repeat ESR; if again positive, corticosteroid therapy
E) Repeat ESR; if normal, additional testing to confirm the diagnosis of TA

My choice was D but A is allegedly correct answer.
I used Likelihood ratio and pretest probability assuming sensitivity is 1(which is almost the same as 99% and doesn't make much difference) to calculate posttest probability which was 5/7.(LR : 2.5, Pretest probability is 0.5 according to question stem). Below is the formula I used.
P' = P0*LR/(1-P0+P0*LR), where P0 is the pre-test probability, P' is the post-test probability, and LR is the likelihood ratio. This formula can be calculated algebraically by combining the steps in the preceding description.(quoted from WIKI 'likelihood ratio')
as P' is equal to positive predictive value, PPV is alreay 5/7. If you do it once again like option D and assume it comes out positive again, then I can calculate 2nd P', which I call P''(double prime) and is 25/29=86% (you use 5/7 instead of 0.5 this time in the position of P0 in formula above). I believe this number is great enough to diagnose Temporal Arteritis and give steroids to prevent blindness with all 'serious adverse effects' because the benefit outweighs the harm here.

Help me!!!
this question is tricky. the post above explains it quite well. so basically to sum it all up, ESR is a screening test which in this vignette was positive. most of the options above are trying to trick you. remember principle of biostats 101. once a screening test is positive (ESR here), you move onto the confirmatory (specific test) which is temporal artery biopsy for TA.
Reply With Quote Quick reply to this message
The above post was thanked by:
CHASER (01-27-2016), zak09 (03-12-2016)
  #7  
Old 01-27-2016
USMLE Forums Newbie
 
Steps History: Not yet
Posts: 7
Threads: 3
Thanked 1 Time in 1 Post
Reputation: 11
Default Thanks all guys.

I am reminded of two principles.

1.Back to basics.

2.Think about what board wants.

Thanks everybody.

Reply With Quote Quick reply to this message
  #8  
Old 03-11-2016
USMLE Forums Veteran
 
Steps History: 1+CK+CS+3
Posts: 274
Threads: 31
Thanked 86 Times in 71 Posts
Reputation: 96
Default

Although answers B through E can easily be eliminated, I am not sure that even A is correct. AS Nodo mentioned, the next step in management is often steroids.

Where is did you find this question?
__________________
Check out my
To view links or images in signatures your post count must be 10 or greater. You currently have 0 posts.
channel on youtube.
Reply With Quote Quick reply to this message



Reply

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
Was CK This Hard Always! drpepper4u USMLE Step 2 CK Forum 78 07-29-2016 02:02 PM
Biostatistics questions Gikson USMLE Step 3 Forum 3 07-04-2015 07:45 PM
Looking for Step 1 Partner in Online/biostatistics !; Exam: Soon ! / biostatistics dede USMLE Step 1 Study Partners 0 03-16-2013 04:49 AM
IM is hard to get into? add1 IMG Residency Match Forum 6 05-08-2011 04:50 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)