[MUSIC] As you know, epidemiologic studies can be split into two types, descriptive and analytic. You've previously learned about descriptive studies and now we're focusing on analytic studies, these deal with individuals. In particular, we're looking at observational studies within the analytic study design group. You've already learned about case control studies, now it's time to look at cohort studies. In relation to the hierarchy of evidence, we're climbing up the ladder. And with regards to observational study designs, cohort studies are considered the most robust. Now I will take you through some of the key features of a cohort study. The cohort study typically involves a group of people without disease who are observed over a period of time to see what happens to them. This is also known as a longitudinal study. As a result, the first step in conducting a cohort study is to select your target population and assess their exposure status. Next you will follow these people to check up if they develop the disease of outcome or outcome of interest. So the defining characteristic of a cohort study is that you track people forward in time, you always assess exposure prior to disease. This is what a typical cohort study looks like. You have a group of healthy individuals at the beginning, you assess them for the exposure of interest. So here the blue circles represent the exposed and the blue squares represent the unexposed. You would then wait for a period of time to pass and then assess the disease status of the individuals. Here, the red circles represent the exposed individuals who went on to develop the disease over time. And the red squares represent the unexposed people who went on to develop the disease. As I've said, you follow the participants up over a period of time. This could be months or it could be years, which depends on your endpoint and for example, the rarity of the disease you're interested in. During the follow up period, you ideally want to keep track of everyone but unfortunately that doesn't always happen. People may die or move away. So your population might be slightly smaller when you finish, depending on the type of people you enroll or the endpoint that you're interested in. At the end of the study, you will have some people who are still healthy, some who have developed the disease of interest, or even developed a different disease and potentially some who have died. So, how do you set up a cohort study? First, let's look at selecting the target population. For example, if you wanted to study a chronic disease that you know is associated with ageing, you would perhaps want to recruit older individuals. Say, those over 50 because you'd be waiting a very long time for a group of 20-year-olds to develop Alzheimer's. It is also important to understand the exposure of interest. So if you're looking at a rare exposure, such as pesticides, you may need to target a specific population, such as farmers. However, you should initially attempt to identify as many subjects as possible without including any restrictions because you want your study findings to be generalizable. The next factor to consider is how the cohort is assembled or identified. This could be by geographic region for example, the Framingham Heart Study or by occupation, as was the case with the British Doctor Study. You could also set up your cohort based on disease, for example the Multicenter AIDS Cohort or by risk group like a San Francisco men's health study, which studied HIV in gay men. Or you could have a birth cohort to follow a group of individuals born in a certain year. These are all just some examples of cohort types. Now let's look at assessing exposures. In a cohort study, multiple exposures are often assessed. The reason for this is because when you get to the analysis phase, you want to be able to control for other exposures, which may be potential sources of confounding. This is a concept you'll learn about in the next course. The exposures must be well-defined. They can be binary, or it could be that all individuals are exposed, but at different levels. For example, obesity could be defined as yes or no, based on a given body mass index threshold. Or it could be categorized into underweight, normal weight, overweight and obese or it could also be a continuous variable. Exposures can be assessed using a variety of methods, including self-report, using for example a questionnaire or taking physical measurements in the clinic. Exposure data can also be collected using existing records such as medical records and census data. So now turning to outcome ascertainment. Outcome measures may also be obtained from various sources including routine surveillance of counts of registry data, death certificates, medical records or directly from the participant. However, the method used to ascertain the outcome must be identical for both the exposed and unexposed groups. To summarize, cohort studies are important observational study designs. Hopefully you now understand some of the key features to consider when conducting a cohort study. [MUSIC]
