question in WORLD about AML and picking the control group in a case control

[aashkab]

---

Members don't see this ad.

Do you guys remember the question in world where a bunch of kids are exposed to some toxic stuff and have AML. They want to do a case control study about it. and the question asks you what the control group should be?

I thought it would be kids who have AML that weren't exposed to the toxic stuff but world is saying that its kids who are exposed to the toxic stuff and DON'T have AML...

am i missing something here or shouldn't that have been the setup for a cohort study . also, apparently only 16% of people got the question right so is it an error?

i don't know how to get the question id, sorry but if you type in AML and case control you may find it in the search tab.

thanks

 

[username456789]

Do you guys remember the question in world where a bunch of kids are exposed to some toxic stuff and have AML. They want to do a case control study about it. and the question asks you what the control group should be?

I thought it would be kids who have AML that weren't exposed to the toxic stuff but world is saying that its kids who are exposed to the toxic stuff and DON'T have AML...

am i missing something here or shouldn't that have been the setup for a cohort study . also, apparently only 16% of people got the question right so is it an error?

i don't know how to get the question id, sorry but if you type in AML and case control you may find it in the search tab.

thanks

In a case control, you're comparing CASES (got the disease of interest) to CONTROLS (by definition, did NOT get the disease of interest, but are otherwise similar to the cases). Then you look to see if a particular exposure/risk factor is present significantly more often in the cases vs. the controls.


Here's an easy example in a smoking case control:

Cases: 50 white middle aged males with lung cancer
Controls: 50 white middle aged males without lung cancer
Exposure of interest: Smoking

Results:
41/50 lung cancer patients had a history of smoking
1/50 patients without lung cancer had a history of smoking

Conclusion: patients with lung cancer have a higher likelihood of also having been smokers.

 

[aashkab]

okay thanks, that helps a lot and was easy. i definitely wasn't getting the big picture with just FA for that.

 

[username456789]

okay thanks, that helps a lot and was easy. i definitely wasn't getting the big picture with just FA for that.

Yeah FA's not great for explaining that kind of thing.

We had a decent biostats/epi section during our second year so some of that stuff stuck with me pretty well.