Introduction to Epidemiology

Full disclosure: I was unable to attend this lecture, but others have told me that all of the necessary content is on the slides, so that's what I'll be using to guide my writing of this post.

Define epidemiology and its role in health

Epidemiology, in a nutshell, is studying how diseases and other health-related states are distributed in a population. This knowledge is very useful in controlling health problems- for instance, if epidemiologists discover that a Hepatitis A outbreak is linked to a batch of berries, the government can respond by alerting everyone who might have bought the berries and/or issuing recalls, so that the outbreak doesn't spread any further.

Define the difference between:

Descriptive and analytical studies

Descriptive epidemiology, as the name suggests, seeks mainly to describe what is happening in terms of characteristics such as person, place, and time. Analytic epidemiology, on the other hand, seeks to go further and answer the how and why questions, and involves the testing of hypotheses.

Observational and experimental studies

In observational studies, epidemiologists just observe what is happening, without intervening in any way. On the other hand, in experimental studies, one or more factors are intentionally altered so that epidemiologists can formally test whether alteration of one of these factors can lead to an increased or decreased risk of disease.

Distinguish between prevalence, incidence of disease and key measures of disease frequency

• Prevalence- The number of cases of the disease within a particular period. Includes both old and new cases. Defined as (All new and pre-existing cases during period) / (Population during same time period).
• Incidence- Similar to prevalence, but only new cases are considered. Defined as (New cases during period) / (Size of population at the start of the period).
• Incidence rate- The rate at which new cases appear. Often expressed as person-time. Defined as (Number of new cases) / (Sum of the time that each person was at risk).

Generally, prevalence is a good metric for general health planning, and for monitoring diseases that don't have a clearly-defined onset. Incidence is a good metric for studying aetiology (the factors leading to development of disease).

Describe the main types of epidemiological study designs

• Cross-sectional- Analyses the population at a particular point in time.
• Prospective cohort- Participants are sorted based on exposed or not exposed to the risk factor of interest. They are tracked over time to see if they develop the disease of interest in the future.
• Retrospective cohort- Participants are sorted based on exposed or not exposed (similar to prospective cohort studies). The number of people who have already developed the disease is calculated. Rates of disease are compared between the two groups.
• Retrospective case-control- Participants are sorted based on whether or not they have the disease. Their histories are then analysed to see if the risk factor of interest is higher in one group than the other.

Note that prospective studies select participants first, and then track them over time to see if a disease develops later. Retrospective studies are the opposite: people already have the disease, and then epidemiologists work backwards from there. Also, note that cohort studies compare rates of disease between exposed and unexposed groups, whereas case-control studies compare rates of exposure between groups with and without the disease.

Define association and interpret key measures of association

An association is a statistical relationship between two variables: for instance, as one variable increases, the other variable might increase as well. Such a relationship does not imply that one variable causes the other. Remember, correlation does not mean causation! There are many factors that could influence statistical relationships other than causality: for instance, there might be a third variable that influences levels of the other two variables (in other words, you could have a confounding factor).

Here are some of the key measures of association used in epidemiology:

• Risk ratio (RR) / Relative risk- Compares the risk of disease or other health event between two groups. Calculated by (Risk of disease in group of primary interest) / (Risk of disease in comparison group).
• Rate ratio- Similar to risk ratio, but compares the rate at which events occur, rather than the risk. Calculated by (Rate for group of primary interest) / (Rate for comparison group).
• Odds ratio- The odds ratio is a little bit more complex in that it also accounts for whether or not people were exposed or not. It can be calculated by (Number of people exposed AND with disease * Number of people unexposed AND without disease) / (Number of people unexposed BUT with disease * Number of people exposed BUT without disease).