4.14: Epidemiological Studies

There are a variety of types of epidemiological studies, studies done in public health. These studies are more specific than the Census or the National Health Surveys. They are studies designed by researchers to answer a specific set of questions. When you are researching for your papers this semester, you will probably come across this type of study. For that reason -- and again, so you can be a critical consumer of information -- we are going to discuss some of the most common types of studies.

Now, this topic is big. You could spend a whole semester on it! And we won't. I just want you to understand some of the main differences among the types of studies.

The first big difference to understand is that between descriptive and analytic studies.

• Descriptive studies describe epidemics with respect to person, place, and time
• Analytic studies are aimed at testing hypotheses

A descriptive study tracks exactly what is happening or has happened in an epidemic or other health phenomenon. It doesn't test one situation against another, systematically. This information is still very useful, and researchers often compare data from one epidemic to data from another epidemic later, in an analytic study. In addition, as a descriptive study tracks data over the course of an outbreak, for example, it can give clues to what the causes of the outbreak are. If you look at the Scenario at the beginning of Chapter 3 in ICPH (p. 67) and again at the end, the Scenario Analysis and Response section (p. 98+), there is an exercise that allows you to take part in analyzing the data from a simple descriptive study, to understand an outbreak of food poisoning.

A descriptive study allows you to plot the data on a graph about who got sick, when and where. This creates an epidemic curve -- a graph of an epidemic. An epidemiologist would know that certain kinds of epidemic curves are typical of certain diseases. Your textbook ICPH gives some examples of these.

For example: As the Covid-19 pandemic first emerged in Wuhan, China, descriptive studies reported on who was getting sick and where. As the pandemic has continued, many other descriptive studies have described how the disease is affecting people -- how many, when, where, who, and with what impacts.

An analytical study is designed to test out ideas (hypotheses) about the relationships between a health problem and a certain cause or risk factor. Risk factors are things that are associated with greater probability of disease (or injury). For example, an analytical study could try to find out whether drinking sweetened soda (possible risk factor) is associated with a higher risk of diabetes (disease).

Analytical studies can be done in several different ways. One approach is observational. The researchers observe a group of people or a community over time and carefully collect data on what happens or on what people report has happened in the past. The researchers don't DO something other than observe.

When the researchers DO something to one group and compare the outcomes with another group, that is called an experimental study. In an experimental study, the groups are separated, either randomly or not. One group gets an exposure of some kind (say, they get an incentive to stop smoking) and the other group does not (no incentive). The researchers look at how the two groups compare.

• Randomization means one group that is randomly selected from the larger population of the study and they are given a treatment or a program that the other group does not receive.
• Placebos are often used when comparing two types of treatment -- one group gets a false kind of treatment (like sugar pills) while the other group gets the medicine, or other intervention.
• Sometimes it is not feasible or ethical to randomize, and instead a control group that is similar to the treatment group is used to compare outcomes.

Observational study example: Thousands of observational analytical studies about Covid-19 are underway. One study could follow the people who report "long Covid" -- symptoms that persist after the recovering from the infection -- to see what happens, tracking different variables they think might have an effect on "long Covid".

Experimental study example: Experimental studies were needed in order to assess whether the new vaccines work. People who volunteered for the clinical trials of vaccines were randomized in a double-blind study (neither they nor the doctors knew who was getting the real vaccine and who was getting a placebo), and both groups were followed to see who would get sick from Covid. The two groups were compared, to see how effective the vaccine was at preventing illness from the coronavirus.