The data includes ample information on medications given in the ED and/or prescribed at discharge. In fact, information on up to 12 medications is allotted in various columns. Obviously, medication administration occurs after the decision to admit the patient has been made, so we cannot use these columns for this use case.

Nevertheless, we encourage you to peruse the documentation and read about the coding systems used for medications if you wish to use such information in your own predictive modeling:

med_cols = [
    'MED1','MED2','MED3','MED4','MED5',
    'MED6','MED7','MED8','MED9','MED10',
    'MED11','MED12','GPMED1','GPMED2','GPMED3',
    'GPMED4','GPMED5','GPMED6','GPMED7','GPMED8',
    'GPMED9','GPMED10','GPMED11','GPMED12','NUMGIV',
    'NUMDIS','NUMMED',
]

X_train.drop(med_cols, axis=1, inplace=True)
X_test.drop(med_cols, axis=1, inplace=True)