Identifying episodes of an illness/outcome

[Carmine_Rossi]

Good afternoon fellow SAS users and communities,

I have a question concerning the identification of episodes of an illness/outcome using data at the level of patient-weeks. Using the hypothetical data below (two patients), every time a binary event occurs, I am interested in flagging/counting the continuous episode of high-cost associated with that event. The variable I am interested in deriving is called "episode". When the high_costs=1 stop, then the episode should stop as well. Once the event=1 starts up again, another episode is occurring and high costs are assessed again. A new episode cannot start until the high_cost returns to 0 (see patient 2). Ultimately, I want to know the episode number of a particular week. 

Would anyone be able to help out with this?

Thanks in advance,

-Carmie

Subject_id	Week	Event	High_cost	Episode
1	1	1	1	1
1	2	0	1	1
1	3	0	1	1
1	4	0	1	1
1	5	0	1	1
1	6	0	1	1
1	7	0	0	0 or deleted
1	8	0	0	0 or deleted
1	9	0	0	0 or deleted
1	10	0	0	0 or deleted
1	11	1	1	2
1	12	0	1	2
1	13	0	1	2
1	14	0	0	0 or deleted
1	15	0	0	0 or deleted
1	16	0	0	0 or deleted
1	17	0	0	0 or deleted
1	18	0	0	0 or deleted
2	1	1	1	1
2	2	0	1	1
2	3	0	1	1
2	4	1	1	1
2	5	0	1	1
2	6	0	0	0 or delete
2	7	0	0	0 or delete
2	8	1	1	2
2	9	0	0	0 or delete
2	10	0	0	0 or delete
2	11	0	0	0 or delete

[Ksharp]

data have;
input Subject_id
Week
Event
High_cost
Episode	;
cards;
1
1
1
1
1
1
2
0
1
1
1
3
0
1
1
1
4
0
1
1
1
5
0
1
1
1
6
0
1
1
1
7
0
0
0 or deleted
1
8
0
0
0 or deleted
1
9
0
0
0 or deleted
1
10
0
0
0 or deleted
1
11
1
1
2
1
12
0
1
2
1
13
0
1
2
1
14
0
0
0 or deleted
1
15
0
0
0 or deleted
1
16
0
0
0 or deleted
1
17
0
0
0 or deleted
1
18
0
0
0 or deleted
2
1
1
1
1
2
2
0
1
1
2
3
0
1
1
2
4
1
1
1
2
5
0
1
1
2
6
0
0
0 or delete
2
7
0
0
0 or delete
2
8
1
1
2
2
9
0
0
0 or delete
2
10
0
0
0 or delete
2
11
0
0
0 or delete
;

data want;
 set have;
 by subject_id;
 if first.subject_id then group=0;
 if high_cost=1 and (lag(high_cost)=0 or first.subject_id) then group+1;
 want=ifn(high_cost=1,group,0);
 drop group;
run;