1
$\begingroup$

we are researching whether the date of a certain medical procedure impacts the complication rate of said procedure. To this end we have identified multiple risk factors we would like to study more closely (e.g. the day of the week, holiday seasons, winter/summer and so on).

Simplified, we have a table of the form

enter image description here

where the reference group is the subset of patients that have neither risk factor A, B or C. Please note that $A \cap B \neq \emptyset$ etc, i.e. we cannot eliminate dependency between columns.

What is now the appropriate statistical test or model to compare the individual risk groups against the reference group? Our null-hypothesis is that group A/B/C do not have more infections in comparison to the reference group.

Improve this question
$\endgroup$

2 Answers 2

Reset to default
1
$\begingroup$

You should look into logistic regression. There are a lot of posts on this site, so search ...

For more specific advice, you will need to give more details.

Improve this answer
$\endgroup$
0
$\begingroup$

You could consider using a linear regression model (if applicable) which allows you to do a ANCOVA test to determine the impact of the risk factors and date on infection rate. $Infections=\beta_0+\beta_1Date+\beta_2Group A+\beta_3GroupB+\beta_4GroupC+\beta_5Reference+\epsilon$

Or you could do a $\chi^2$ test of homogeneity between the different treatment groups.

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.

Not the answer you're looking for? Browse other questions tagged or ask your own question.