Hello,
I have a dataset with all values for some variables (clinical characteristics of the patient) and missing values for other variables (socioeconomic characteristics of the patient).
I want to use both of these variables in a regression model so as to explain a medical intervention.
My variable of interest is the one with missing values (socioeconomic variables).
My objective is to study the effet of socioeconomic conditions on a medical intervention, controlling by medical variables.
I know that i can use Multiple Imputation by Chained Equations to handle missing values but i wonder if i can use it when missing values concern only control variables or also when it concerns my variables of interest.
In other words, i wonder if i can interpret the results of my variables of interest using Multiple Imputation by Chained Equations to handle missing values or i should only use availaible observations of my variable of interest.
Any help?
Thanks in advance.