EPIDEMIOLOGY Flashcards
What is a case control trial?
A case control study is an example of an observational study. What this means is that we don’t actually do anything to the participants (like give them a drug) we just observe them. In a case control study you get a group of people with the disease or condition that you are interested in and a group of people without the condition (the controls). Then you find out about these different groups and try and work out any differences between them that might explain why those that developed the condition did so.
What is a cohort study?
A cohort study is a type of observational study. What this means is that we don’t actually do anything to the participants of the study (like give them a drug) rather we just observe them. A cohort study involves getting a group of people with particular risk factor that you are interested in (eg smoking) and a group of people without the risk factor and watching the different group over a set period of time. You can then see which of them develops a disease that you are interested in such as lung cancer.
What is a randomised control trial?
A randomised control trial is a type of experimental study and is actually the gold standard (the best) when it comes to types of studies and how good they are at proving something. A RCT involves collecting participants and randomly assigning them into either group A or group B. You then do different things to each group. For example you could give group A a drug that was being tested and group B a placebo. You then compare the two groups to see if any differences appear. The reason RCTs are so good is that firstly because you have randomised the groups you can be pretty happy that they are very similar and therefore you can also be pretty happy that any difference between the groups is due to the factor that you gave them. You also make sure that the participants don’t know which group they were assigned to so that any differences are truly because of the organic effect of the drug rather than because the patient thinks they have been giving the drug. This is called blinding.
What is a cross sectional study?
A cross sectional study is an example of an observational study. What this means is that we don’t actually do anything to the participants (like give them a drug) we just observe them. A cross sectional study most commonly involves getting people to fill in a survey that asks a variety of different questions and then trying to find trends within the collated data. The data is therefore collected at only one point in time. It is good at picking broad correlations such as salt and heart disease because it can easily collect data from a huge number of people. The problem with this form of study is that it can’t go into enough detail to work out if there any confounding factors that might be responsible for the trend.
What is the p-value?
The p value tells us likelihood that the pattern that you’re study identified is due to chance alone. So say you are comparing two groups and you find a difference between them, you then want to know whether the difference you have found is actually representative of the wider population and is therefore significant or whether it was just due to chance and actually it cannot be applied to the wider population. The cut off that we tend to choose is normally 0.05. If the p value is less than 0.05 then that is saying that the likelihood that you’re result is not significant is less than 5%.
What is the relative risk?
The relative risk is the likelihood of an event happening with a particular factor versus the likelihood of the same event happening without the factor. So say the event that we are looking at is having a heart attack and the factor we are testing is a particular drug. We have two groups, group A taking the drug and group B not taking the drug. If when we compare the groups the relative risk of taking the drug is less than 1, then there is a deceased risk of having a heart by taking the drug in comparison to not taking the drug. Equally if the relative risk of taking the drug is above 1, then there is an increased risk of having a heart by taking the drug in comparison to not taking it. If the relative risk is 1 then the likelihood of having a heart attack is the same whether you are taking the drug or not.
What is the confidence interval?
When we do a study it is to try and apply any patterns that we have found in our sample to the general population. Say our study is comparing any outcomes of taking a drug with not taking a drug. Any difference that we find from our sample will never be exactly the true difference that exists in the wider population of people taking the drug versus those not taking it. So the confidence interval allows for some error and gives us the range within which we are 95% sure that the real value lies. If the range is small the we can be pretty happy that the difference seen in our sample is a good estimate of the true difference seen at a population level. However, if the range is wide then the true degree to which the drug has an effect cannot be precisely estimated.
