# How to use case_when() for exclusive multiple conditions

Solution for How to use case_when() for exclusive multiple conditions
is Given Below:

In our study, a participant is regarded as a case with hypertension when he/she has a history of hypertension, or systolic blood pressure >130 mmHg, or diastolic blood pressure >80 mmHg, or recieves antihypertensive treatment. I create the final hypertension status by using two different methods as listed below, and the results of method 1 is correct but not method 2. My question is how to create the final hypertension status by using the case_when() function in method 2?

``````#-------------------------- data set
# id: id number of participants
# hptn_his: self-reported history of hypertension (0 means no, 1 or 2 mean yes)
# sbp: systolic blood pressure (mmHg)
# dbp: diastolic blood pressure (mmHg)
# treat: whether recieved the antihypertensive treatment (0 means no, 1 means yes)
id<-c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
hptn_his<-c(0, NA, NA, NA, NA, NA, 2, 2, 2, 0, 0, NA, NA, 0, 0, 0, 0, 0, 0, NA, 0, 1, 1, 1)
sbp<-c(NA, NA, NA, NA, NA, 110, 105, 115, NA, NA, NA, NA, 109, 102, 140, 136, 150, 126, 112, 147, NA, NA, 155, 124)
dbp<-c(NA, NA, NA, NA, NA, 61, 62, 84, NA, NA, NA, NA, 75, 67, 74, 67, 58, 50, 45, 48, NA, NA, 80, 74)
treat<-c(NA, NA, NA, 1, NA, NA, NA, 1, NA, NA, 1, NA, NA, NA, NA, NA, NA, NA, NA, 1, 1, NA, 1, NA)

mydata<-as.data.frame(cbind(id,hptn_his,sbp,dbp,treat))

mydata

id hptn_his sbp dbp treat
1   1        0  NA  NA    NA
2   2       NA  NA  NA    NA
3   3       NA  NA  NA    NA
4   4       NA  NA  NA     1
5   5       NA  NA  NA    NA
6   6       NA 110  61    NA
7   7        2 105  62    NA
8   8        2 115  84     1
9   9        2  NA  NA    NA
10 10        0  NA  NA    NA
11 11        0  NA  NA     1
12 12       NA  NA  NA    NA
13 13       NA 109  75    NA
14 14        0 102  67    NA
15 15        0 140  74    NA
16 16        0 136  67    NA
17 17        0 150  58    NA
18 18        0 126  50    NA
19 19        0 112  45    NA
20 20       NA 147  48     1
21 21        0  NA  NA     1
22 22        1  NA  NA    NA
23 23        1 155  80     1
24 24        1 124  74    NA

#-------------------------- method 1, correct
mydata<-within(mydata,{
hptn1<-NA
hptn1[hptn_his==0]<-0
hptn1[hptn_his==1|hptn_his==2|sbp>130|dbp>80|treat==1]<-1
})

table(mydata\$hptn1)

0  1
5 13

#-------------------------- method 2, incorrect
library("dplyr")
mydata<-mydata%>%
mutate(
hptn2=case_when(hptn_his==0~0,
hptn_his==1|hptn_his==2|sbp>130|dbp>80|treat==1~1,
TRUE~NA_real_))

table(mydata\$hptn2)

0  1
10  8

#-------------------------- comparison
# hptn1: hypertension status by method 1 (0 means no, 1 means yes), correct
# hptn2: hypertension status by method 2 (0 means no, 1 means yes), incorrect
mydata

id hptn_his sbp dbp treat hptn1 hptn2
1   1        0  NA  NA    NA     0     0
2   2       NA  NA  NA    NA    NA    NA
3   3       NA  NA  NA    NA    NA    NA
4   4       NA  NA  NA     1     1     1
5   5       NA  NA  NA    NA    NA    NA
6   6       NA 110  61    NA    NA    NA
7   7        2 105  62    NA     1     1
8   8        2 115  84     1     1     1
9   9        2  NA  NA    NA     1     1
10 10        0  NA  NA    NA     0     0
11 11        0  NA  NA     1     1     0
12 12       NA  NA  NA    NA    NA    NA
13 13       NA 109  75    NA    NA    NA
14 14        0 102  67    NA     0     0
15 15        0 140  74    NA     1     0
16 16        0 136  67    NA     1     0
17 17        0 150  58    NA     1     0
18 18        0 126  50    NA     0     0
19 19        0 112  45    NA     0     0
20 20       NA 147  48     1     1     1
21 21        0  NA  NA     1     1     0
22 22        1  NA  NA    NA     1     1
23 23        1 155  80     1     1     1
24 24        1 124  74    NA     1     1
``````

`case_when` returns on the first value that is true. For the records that appear wrong using your second method, these occur when `hptn_hist==0` triggers first before the additional question.

``````mydata<-mydata%>%
mutate(
hptn3=case_when(hptn_his==1|hptn_his==2|sbp>130|dbp>80|treat==1~1,
hptn_his==0~0,
TRUE~NA_real_))
``````

More detailed explanation:

• In your first approach you are using over-writing. Some values from `hptn1[hptn_his==0]<-0` are then overwritten by `hptn1[hptn_his==1|hptn_his==2|sbp>130|dbp>80|treat==1]<-1`. This is the correct order if later values overwrite earlier onces.

• In your second approach, `case_when` returns only the first true value. So later values can not overwrite earlier ones. Hence the correct order is `hptn_his==1|hptn_his==2|sbp>130|dbp>80|treat==1` before `hptn_his==0`.

Another was to think of this:

``````case_when(condition1 ~ 1,
condition2 ~ 2,
condition3 ~ 3,
...)
``````

Is equivalent to:

``````case_when(condition1 ~ 1,
!condition1 & condition2 ~ 2,
!condition1 & !condition2 & condition3 ~ 3,
...)
``````