Confounding Bias versus Effect Modification! - 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 02-11-2012
avash.nrs's Avatar
USMLE Forums Guru
 
Steps History: Step 1 Only
Posts: 441
Threads: 53
Thanked 258 Times in 153 Posts
Reputation: 268
Stats Confounding Bias versus Effect Modification!

please can anyone tell me the difference between confounding bias and effect modification..well its explained in uworld but could not get it..both appears same to me.
please can u explain this stuf..
Reply With Quote Quick reply to this message
The above post was thanked by:
drzamzam (08-17-2015), sadaf1987 (08-24-2015)



  #2  
Old 02-12-2012
USMLE Forums Addict
 
Steps History: 1+CK+CS
Posts: 134
Threads: 21
Thanked 70 Times in 42 Posts
Reputation: 80
Default

I am too confused on that
Reply With Quote Quick reply to this message
  #3  
Old 02-12-2012
USMLE Forums Scout
 
Steps History: ---
Posts: 58
Threads: 1
Thanked 141 Times in 19 Posts
Reputation: 151
Default

Effect modification is about stratification.

Means that there's difference among the different subgroups of the population under study.

For example;
Drug X worked on children but did not work on older adults. This is effect modification.

Confounding bias on the other hand, is for the entire population under study. Means you cannot say drug X worked or not because age was a confounding factor in the example above.

I know it's so confusing but just try to think this way, (subgroups means effect modification, entire study means confounding).

If you want to really take the headache to understand that, read this article
http://courses.washington.edu/b570/h...onfounding.pdf
Reply With Quote Quick reply to this message
The above post was thanked by:
avash.nrs (02-12-2012), wa87el (01-29-2015)
 
  #4  
Old 02-12-2012
d_wiqed's Avatar
USMLE Forums Master
 
Steps History: Step 1 Only
Posts: 861
Threads: 19
Thanked 891 Times in 432 Posts
Reputation: 908
Default

From Pre-Test Preventive Medicine and Public Health (a usmle step 2 reviewer):

Confounding:
- Is a "nuisance" factor that needs to be eliminated
- It causes a distortion of the results, simply because that factor is distributed unevenly in exposed and unexposed individuals.

For example:
- The association between coffee and lung cancer may be distorted by smoking. Because smoking is unevenly distributed between the two groups.

So when coffee and lung cancer are combined, it seems there is an association. But when smokers and non-smokers are considered separately for each group, there is actually no association between coffee and lung cancer.

Effect modification
- Is not a nuisance factor. It actually provides important information.
- The magnitude of the effect of an exposure on the outcome will vary according to the presence of a third factor.

In the example above:
- In contrast to confounding, effect modification does not depend on the fact that one group has "more smokers" than the other group.

Another example of effect modification:
- If a baby has malnutrition, a high bilirubin seems to be a much stronger risk factor for bilirubin-induced brain damage.

- Malnutrition here is not a confounding factor. Malnutrition is not unevenly distributed between, for example, "brain-damaged" and "no brain damage" groups. As a third factor, it increases the impact of high-bilirubin in relation to bilirubin-induced brain damage.

IN A NUTSHELL:

Confounding factor
The 3rd factor (i.e. smoking) is unevenly distributed between groups A (i.e. lung cancer) and B (i.e. coffee), making groups A and B seem related but they are not.

Effect modification
The 3rd factor (i.e. malnutrition) increases the impact of an exposure (i.e. high bilirubin) on an outcome (i.e. brain damage).


It's a bit hard to explain but I do hope this helped!

@avash.nrs: If it DID help and we meet someday, you owe me a... COKE!!
Reply With Quote Quick reply to this message
The above post was thanked by:
Ahmed127 (11-19-2012), Aimy (08-23-2015), ak-06 (10-02-2016), alice (10-11-2012), anusha reddy (01-31-2015), avash.nrs (02-12-2012), chriledi (02-27-2013), dr_serge (10-30-2014), Dr. Mexito (04-07-2012), Dr.NickRiviera (05-31-2012), drishin (11-17-2014), fatboy1_17 (04-02-2015), Heli kapadia (11-19-2015), Infy (04-12-2016), Janakpriya (06-15-2013), koolkiller88 (09-25-2012), Master shifu (02-12-2012), menaabdonasry (11-26-2014), MishkaMD (09-16-2013), Mr.wonderful (01-29-2015), navu (05-31-2012), Nijhum (11-10-2014), phlebgirl (06-10-2014), sm3356 (07-09-2015), steps_8186 (03-31-2012), Taayseer (10-08-2014), voicesinmyhead (09-08-2014), zainsyed (10-20-2015)
  #5  
Old 02-12-2012
avash.nrs's Avatar
USMLE Forums Guru
 
Steps History: Step 1 Only
Posts: 441
Threads: 53
Thanked 258 Times in 153 Posts
Reputation: 268
Default

Quote:
Originally Posted by d_wiqed View Post
From Pre-Test Preventive Medicine and Public Health (a usmle step 2 reviewer):

Confounding:
- Is a "nuisance" factor that needs to be eliminated
- It causes a distortion of the results, simply because that factor is distributed unevenly in exposed and unexposed individuals.

For example:
- The association between coffee and lung cancer may be distorted by smoking. Because smoking is unevenly distributed between the two groups.

So when coffee and lung cancer are combined, it seems there is an association. But when smokers and non-smokers are considered separately for each group, there is actually no association between coffee and lung cancer.

Effect modification
- Is not a nuisance factor. It actually provides important information.
- The magnitude of the effect of an exposure on the outcome will vary according to the presence of a third factor.

In the example above:
- In contrast to confounding, effect modification does not depend on the fact that one group has "more smokers" than the other group.

Another example of effect modification:
- If a baby has malnutrition, a high bilirubin seems to be a much stronger risk factor for bilirubin-induced brain damage.

- Malnutrition here is not a confounding factor. Malnutrition is not unevenly distributed between, for example, "brain-damaged" and "no brain damage" groups. As a third factor, it increases the impact of high-bilirubin in relation to bilirubin-induced brain damage.

IN A NUTSHELL:

Confounding factor
The 3rd factor (i.e. smoking) is unevenly distributed between groups A (i.e. lung cancer) and B (i.e. coffee), making groups A and B seem related but they are not.

Effect modification
The 3rd factor (i.e. malnutrition) increases the impact of an exposure (i.e. high bilirubin) on an outcome (i.e. brain damage).


It's a bit hard to explain but I do hope this helped!

@avash.nrs: If it DID help and we meet someday, you owe me a... COKE!!
GOT it sumwhat in my 'tiny' brain.these things really irritate me..such subtle differences,not even in kaplan and fa.Uworld is absolutely..i just pray the real xam is not exactly like uworld in total 100%......that wud be disstrous..so much hard qs to answeras far as the coke is concerned yeah mate,no problem..a MAC D burger and coke will be great..hope we meet and meet real soon..its been a gr8 experience talking to u and xapezx and am just amazed by the help u people did to methanks for everything
Reply With Quote Quick reply to this message
The above post was thanked by:
d_wiqed (02-12-2012)
  #6  
Old 02-12-2012
d_wiqed's Avatar
USMLE Forums Master
 
Steps History: Step 1 Only
Posts: 861
Threads: 19
Thanked 891 Times in 432 Posts
Reputation: 908
Default

Quote:
Originally Posted by avash.nrs View Post
GOT it sumwhat in my 'tiny' brain.these things really irritate me..such subtle differences,not even in kaplan and fa.Uworld is absolutely..i just pray the real xam is not exactly like uworld in total 100%......that wud be disstrous..so much hard qs to answeras far as the coke is concerned yeah mate,no problem..a MAC D burger and coke will be great..hope we meet and meet real soon..its been a gr8 experience talking to u and xapezx and am just amazed by the help u people did to methanks for everything
I was also thinking of a MAC D! But I was a bit shy...

But now that you have mentioned it...

Can you make it a CHEESEburger with fries? Thankee!! Lol!
Reply With Quote Quick reply to this message
The above post was thanked by:
Hitman (02-12-2012), Nijhum (11-10-2014)
  #7  
Old 02-12-2012
avash.nrs's Avatar
USMLE Forums Guru
 
Steps History: Step 1 Only
Posts: 441
Threads: 53
Thanked 258 Times in 153 Posts
Reputation: 268
Default

yeah of course..anything for a friend
Reply With Quote Quick reply to this message
The above post was thanked by:
d_wiqed (02-12-2012), Nijhum (11-10-2014)
  #8  
Old 02-20-2012
dr-ahmed's Avatar
USMLE Forums Guru
 
Steps History: 1+CK+CS
Posts: 333
Threads: 61
Thanked 390 Times in 90 Posts
Reputation: 400
Default One of the best explanations I have ever read about the dilemma of confounding vs effect modification

The difference between confounding , effect modification in bio statistics :

1- Confounding bias :

Alcohol ( exposure ) Oral cancer ( outcome )

Smoking is a confounding bias here. Why ?

Because smoking is related to BOTH the exposure and the outcome. People who smoke will more often than not, also drink, and people who have oral cancer, were probably smokers ( smoking is a known risk factor for oral cancer).

2- Effect modification :

OCPs ( exposure ) Breast cancer (outcome )

Family history of breast cancer is an effect modification here.
Can you guess why ?

Asbestos ( exposure ) Lung cancer(outcome)

Smoking is an effect modification

Estrogens ( exposure ) DVT (outcome )

Smoking is an effect modification

So do you notice the difference ?? The effect modification is ONLY related to the outcome, but NOT to the exposure. Think with me here : smoking does not effect neither asbestos exposure nor estrogen levels or intake,BUT definitely is a risk for lung cancer and DVT !!

Family history of breast cancer has nothing to do with choosing to use OCPs , but definitely has a factor in increasing risk of breast cancer !!
Reply With Quote Quick reply to this message
The above post was thanked by:
acr2503 (08-21-2012), Ahmed127 (11-19-2012), ak-06 (10-02-2016), anamta (04-26-2016), anusha reddy (01-31-2015), avash.nrs (03-30-2012), Aya101 (11-11-2016), blackychen (09-22-2015), carla (06-18-2013), chriledi (02-27-2013), datson (09-01-2015), dockhi (06-01-2012), doxorubicin (02-20-2012), dr_kg (04-13-2016), Dr. Mexito (04-07-2012), drhma (05-31-2012), drishin (11-17-2014), drnrpatel (02-20-2012), drsrb (11-03-2013), erickven (09-22-2016), gargabhi2 (02-20-2012), hope2015 (06-24-2014), Infy (04-12-2016), Janakpriya (06-14-2013), jeniaz (12-04-2013), koolkiller88 (09-25-2012), LEE WON HYUK (05-22-2013), mle_vrp (10-27-2015), mohanned (05-22-2015), Mr.wonderful (01-29-2015), navu (05-31-2012), Neutrino (06-17-2014), Nijhum (11-10-2014), Oscaryasser (10-21-2013), phlebgirl (06-10-2014), sadaf1987 (08-24-2015), steps_8186 (03-31-2012), Taayseer (10-08-2014), voicesinmyhead (09-08-2014)
  #9  
Old 02-20-2012
USMLE Forums Addict
 
Steps History: 1+CK+CS
Posts: 134
Threads: 21
Thanked 70 Times in 42 Posts
Reputation: 80
Default

Marvellous
Reply With Quote Quick reply to this message
  #10  
Old 02-20-2012
dr-ahmed's Avatar
USMLE Forums Guru
 
Steps History: 1+CK+CS
Posts: 333
Threads: 61
Thanked 390 Times in 90 Posts
Reputation: 400
Default copied

This is copied by the way
Reply With Quote Quick reply to this message
  #11  
Old 02-20-2012
USMLE Forums Scout
 
Steps History: 1 + CS
Posts: 59
Threads: 12
Thanked 10 Times in 10 Posts
Reputation: 20
Default

super cool....
Reply With Quote Quick reply to this message
  #12  
Old 08-19-2014
USMLE Forums Newbie
 
Steps History: CK Only
Posts: 1
Threads: 0
Thanked 5 Times in 1 Post
Reputation: 15
Thumbs Up Simple&practical

The simple and practical strategy to differentiate the Confounding from Effect Modification is to stratify the results based on the new factor(as most questions do it):
1-In CONFOUNDING; the relation in new subgroups is no longer meaningful and taking the confounder into account will distort the primary positive relation:
Example:
Alcohol------->Bladder Cancer RR=2 p-value<0.05
Now after stratification:
Alcohol in non-smoker---------->Bladder Cancer p>0.05
Alcohol in smoker----------------->Bladder Cancer p>0.05
2- In Effect Modification; the stratification based on new factor make the primary result more strong in one subgroup
Example is to categorizing patients based on certain HLA, Alleles, Ages, Family History
Alanine dye exposure------> bladder cancer RR:2 p<0.05
Now after stratification:
Alanine dye and positive Family Hx----> Bladder Ca RR:3 p<0.05
Alanine dye and no family Hx--------> Bladder Ca RR:1.5 p<0.05
Reply With Quote Quick reply to this message
The above post was thanked by:
doc5428 (10-13-2014), dr_serge (10-30-2014), Mr.wonderful (01-29-2015), Neutrino (10-15-2014), Taayseer (10-08-2014)
  #13  
Old 01-29-2015
USMLE Forums Scout
 
Steps History: Step 1 Only
Posts: 56
Threads: 5
Thanked 41 Times in 20 Posts
Reputation: 51
Default

You guys are wonderful!
Was going to skip this as uworld didn't explain it as such clearly!
Thanks

Last edited by Mr.wonderful; 01-29-2015 at 06:25 AM.
Reply With Quote Quick reply to this message



Reply

Tags
Biostatistics-Epidemiology

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
The difference between Effect modification and Confounding bias Moonti USMLE Step 2 CK Bits & Pieces 3 01-03-2015 02:48 PM
Confounding bias versus Effects Modification tinhi USMLE Step 1 Forum 9 10-26-2014 10:14 AM
The difference between Effect modification and Confounding bias Moonti USMLE Step 2 CK Forum 0 01-05-2012 09:16 AM
Differenece between Confounding & Effect Modification dr-ahmed USMLE Step 2 CK Forum 4 12-24-2011 02:24 PM
Biostatistics #9 - Confounding, effect modification and adjusted analysis bebix USMLE Step 1 Forum 18 06-06-2011 05:25 AM

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)