2
$\begingroup$

My current issue lies within EMR extracted data for medications. There are multiple variables named: Medication_1, Medication_2, Medication_3, etc...

Refer to attached picture

This data may overlap and analyzing each column seems meaningless as each row takes all these medications at the same time.

Is it better to analyze these drugs in clusters? or are there other ways to deal with this data? In other terms, what is the best way to showcase such data?

Improve this question
$\endgroup$
1
  • $\begingroup$ Try organizing your data with columns corresponding to each drug (there are 10 in your example), and the data in teach column being 1 if a subject took that drug, 0 otherwise. It should start making more sense. You can then easily analyze which drugs are most often/rarely prescribed, which are correlated (prescribed together), etc. BUT; it depends on what you are trying to find out about this data? Can you clarify your objective? What questions are you trying to find answers to? $\endgroup$ Commented Jun 29, 2024 at 19:26

1 Answer 1

Reset to default
1
$\begingroup$

Your data structure is called categorical multiple response set (MRC). Each variable is "next answer" and the values are the responses (or, better, numeric response codes).

Another structure to store the same data is dichotomous multiple response set (MRD). Each variable is the response variant (the medication, in your case) and values are binary - selected / true / yes (1) and not selected / false / no (0).

Both ways to hold multiple response data are used. MRC is not suited for some analyses, for example, using the variables as predictors in a regression. MRD is not parsimonious: if categories (response variants) are many, many columns (variables) are needed.

MRC is able to store multiple instances of the same category (such as potential answers to question "gadgets of which manufacturers do you own?").

Any good statistical software must provide:

  1. A way to analyze (tabulate) both structures.
  2. A way to quickly convert one structure into the other.

So, consult with the documentation or manuals of the software you use.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.