Owen Yang
Here is a list of core competences in epidemiology. I am thinking of finding a time to cover them one by one.
Counts, rates, and relative rates
These are about how to quantify number of events in terms of counts, rates and odds (prevalence and incidence, sensitivity and specificity etc.), or different types of relative rates such as rate ratios, odds ratios, incident rate ratios, hazard ratios, or variations such as absolute risk differences.
In all the rates there is an important concept about what the denominators should be. For example, the incident rates should be the events that arise from eligible person-time in which event are theoretical possible, depending on what the research question is.
Prototypes of study design with intention to understand causation
I am not a fan of jargons, but here we are talking about trials, cohort studies, case-control studies, and cross-sectional designs.
At the time when this article is written, there is a popular view to point out whether it is prospective or retrospective cohort studies, or whether there is a case-cohort study design or a nested case-control study design. These are good study designs and it is good to know and to expect to see them, but I am not keen to believe these jargons will be as timeless as the four prototypes.
There are study designs that are not about causations, for example descriptive studies.
Confounding, mediation, effect modification, errors and biases
I feel the concept of them are more important than remembering different types of jargons such as selection bias or immortal time bias etc.
As a student it is very confusing because given these biases and errors why do we not just conduct a randomised control trial for everything. Or what is the point if someone would conduct a study with so many biases. As we gain experiences in research and in the topic, we will start to have our opinion which biases are more important and need addressing.
Sometimes I ask a student try to list 30-40 biases and see whether they can group them into categories. I also ask them to force themselves to point out strengths and weaknesses in a study but limited to 1-3 most important ones.
