Medical statistics and Data Science: Statistics

Tips and tricks in Stata

Content

  1. How to combine several binary variables: scenario 1

How to combine several binary variables: scenario 1

It is common that we combine several binary variables to be a single variable. For example in medical field, we combine many diseases to be a single variable indicating comorbidity score, which can be further categorized any way that one wishes (for example Charlson Comorbidity Index).

Three steps:

  1. Dichotomize all diseases as binary variables, which should be coded as 1 or 0, for example disease1 disease2 disease3
  2. Key step, combine all the dichotomized diseases variables, stata codes: egen comorbidity=rowtotal(disease1 disease2 disease3)
  3. Categorize the generated variable "comorbidity" as you want