Using Odds ratio in case control studies

[@olarcharles9149]

I am in my first year and my first course unit is Epidemiology, I am benefiting from this lecture, I am still grasping the whole concept. Thank you so much

[@phanicemaroria9177]

well elaborated hv been given an assignment have used all your example to work it out thanks alot

[@nurhafizsahzaimey8433]

hi! I just wanna thank you for this video. I just finished my epidemiology paper today and I couldn't do it if its not because of you. thank you so much for your videos. you're the real MVP. =)

[@musaabass5731]

Well explained. Thanks 👍

[@FRACP]

This is excellent! Other videos I found simply went through the 'cross-method' without explaining the proper way first (which I figured out myself) but I'm glad you're video confirmed it!

[@evabudinska692]

Hi Elizabeth, thanks for this video. I was wondering which of the two types of interpretations of OR is most common. The second one is to calculate OR as (a/b) / (c/d) which gives the same result: (a*d)/(c*b), however, it is interpreted as odds ratio between odds of event in the exposed group vs odds of event in the non-exposed group.

[@Daniel-rk2qz]

+Elizabeth Lynch So... can we say that we only use case-controls when a disease (outcome) is rare and so we want to find the risk factors? Whereas otherwise, if the disease is prevalent, then it is a BETTER option to use a cohort study? And does this also mean that our ULTIMATE GOAL is to find a good estimation for a relative risk in order to determine how it affects the person's pre-test probability? Thank you! this is probably the most challenging part of med school haha.

[@shinom0ri]

Pobably the best video to illustrate OR I've seen. For visualisation purposes I think showing why OR approaches RR with an example might be a good idea, since it's not hard to demonstrate, but seems like a neubulous statement often mentioned otherwise...

[@aishasultan8003]

Thank you so much, such a helpful and simple video.

[@somebody6492]

Hello do you have a vid on how to calculate unmatched ratio? Ty.

[@zeedee6688]

great explanation. Thanks

[@Neunhoffer]

Hi. I am trying to figure out how the cdc arrived at 2.34 times odds for their covid study in kentucky. I cant figure out how this is calculated. Can you help?

[@spencerwilliams3385]

Why do we cross multiple? Thanks for the video.

[@RamirezHD]

I was looking on how to interpret the null hypothesis for the odds ratio...but nice video.

[@ElizabethLynch]

Not sure what you mean. The null hypothesis would be an OR of 1.0, meaning there was an equal likelihood of seeing the exposure in cases and controls.

[@DPREY1234]

So, is it correct to say that "I can use OR in cohort studies only when the prevalence is very low (rare disease)" ?????

[@Wald4267]

Because the outcome is rare does that mean 1.11 is a big increase in risk

[@geosui3522]

Very helpful video..I have a question though, even though it has not been mentioned in this video: how we calculate the logarithm of Odds Ratio? for example, if the odds ratio is 0.71, how do we calculate InOR?

[@crystal5263]

How can you tell the disease is rare or not rare if all that is given is the Odds Ratio is 5 (p > .001)?

[@ElizabethLynch]

This is not something you can determine from the OR. you need information about the prevalence of the disease.

[@crystal5263]

Elizabeth Lynch We were not given the prevalence of the disease the question is " Case control study reviewing relationship between Tampon use and Toxic shock syndrome (TSS). OR comparing the two was 5.0 (p < 0.001). They conclude risk of TSS is 5 times higher in tampon users. Conclusion is valid if which following assumption is satisfied? Then the answer was TSS is a rare disease thus a low prevalence... So what I was wanting to know is how did they come to conclude it had a lower prevalence from that question stem?

[@ElizabethLynch]

Crystal Hinojosa Because the OR should only be used to estimate relative risk when the outcome is rare, and they are using the OR to estimate relative risk.

[@crystal5263]

I'm not understanding what you mean by the outcome is rare? How do I know it is a rare disease is there a certain target number? Since it was only 5 times higher does that mean rare?

[@RahulKumar-jk6oi]

Crystal Hinojosa Can you please provide the reference you looked for the outcome? However, it is good approximation of the equivalence between odd ratio to relative risk ratio when prevalence of disease is

[@haarisarain5048]

watching this in 2020 feels very relevant.

[@CraigDeLarge]

+elizabeth, Thanks for these. You are becoming my salvation in a course I am taking now. What markup device and software are you using in this video? I want to adopt it for work I am doing. Thanks for reply.

[@peenleib]

Still don't understand why you can't use RR in case-control studies (in fact you have all numbers needed to calculate it). Could you explain it for me once again?

[@ProfFeinman]

Risk = probability = wins / total events. Odds = wins / losses = wins / (total events - wins). For example: Risk of drawing an ace from fair deck = 4/52 = 1/13 = 0.077 Odds = 4/48 =1/12 = 0.083. If the comparison group also has low incidence, the ratios will also be similar. So, as she says on the video, if even is rare, odds and risk are similar and RR and OR may be the same.

[@maxamedcabdulilaah3299]

Thank u teacher

[@jinikaty1908]

Thanks

[@muhakutv9503]

Thank you so much

[@kenthhuncho6887]

U helped me coz i dodged this lecture

[@dr.satyabratasahoo5644]

Grand

[@mohammed.f7773]

thank you so much