3:40: "...you can definitely tell causality because exposure precedes disease." What about confounding then, which is a problem for both cohort and case-control?
In retrospective cohort, the exposures are already known (so are the outcomes) before the study starts. The group of people that are part of the study are chosen based on the exposures they have encountered. In case-control, exposures are unknown when the study start. It means that you pick up a group of sick people and another one of people who are not. The group studies are chosen based on the fact that they have the disease regardless of their past exposure. The study is then looking backward on these exposures to look for some that might be related to the disease. When analysing the potential relationship of an exposure (smoking) to a disease (cancer) you can evaluate the (exposed and sick; exposed and not sick; not exposed and sick; not exposed and not sick). And then calculate the chance that is truly having an effect to getting the disease. With retrospective cohort study, it's easier to find the 4 conditions when building your group study since you know the exposures of all the patients. With case control it can be more tricky. For instance you could have only smoking people amongst both group of selected people (sick and not sick). Because you didn't know their exposure before. Of course it's easy to ask them if they are smoking but this exemple has to be taken as if you don't know smoking might be a cause. So it's tricky because you kinda gamble with the exposures of your selected study groups. That's how I see it. I might be wrong but I hope it can help somes getting the idea!
@@S.sugar799 In retrospective cohort, you can study multiple diseases that you suspect may be caused by a particular risk factor (exposure). In case control study, you can study multiple risk factors that you suspect may be cause of a particular disease.
In retrospective cohort, all the patients have the "exposure" and you try to discover some relationship with the outcome. In case-control, not everyone has the exposure.
Thank you! I use Sketchbook Express (available free on the app store) as the drawing program, and Camtasia as the screen capture and video editing software. I hope that helps.
I have a science fair coming up at my high school and I am really interested in doing a Case-Control Study. I have two questions, first, I wanted to do a study to see if suffering from a traumatic injury that resulted in the fracture increases a person chance of developing Rheumatoid Arthritis. Would this be a good study? and Also about finding Cases and Controls how would suggest that I set out to do this?
Hears a 6 year late answer. Easiest way would be having access to a hospital/clinic's database in order to extract this kind of information. So only if a family friend/relative who's in the medical field can definitely hook you up. Other than that you can do a systematic review/meta-analysis on case reports/case series using Pubmed.gov and to extract an overall odds ratio (OR) for the risk factors your analyzing. If you do that my guy, your female teachers/classmates will be all over ur cock. Works just as well in medical school, trust me
What type of study design and measure of association would you suggest to study the possible association between AstraZeneca vaccine against SARS- CoV-2 and blood clot? Is case-control applicable?
What type of study design was used? A study designed to identify risk factors associated with myocardial infarction (MI) was initiated in 1990 and is still underway. All of the 1000 men randomly selected to participate in the study in 1990 were free of CHD. Two hundred of these subjects were classified as hypertensive while the remaining subjects were normotensive. At the 20 year follow-up in 2010, 50 of the 200 hypertensive subjects had suffered an acute MI. Eighty of the normotensive subjects had suffered an MI during this period (1990-2010).
You tell us we cannot confirm the casuality in case control bcz whether or not this specific exposure caused for disease. Thats okey But how can we confirm casuality for sure in cohort ?
