Grouping subject visit and assign dates

[tsureshinvites]

I am not aware how to gourp the observations.

I am new to clinical SAS, need help of for the below queries.

Raw Data;

input id$ Visit $ Date $

datalines;

Raw Data
ID	Vist	Start Date
B1110001	Baseline	4/18/2016
B1110001	Visit	4/19/2016
B1110003	Baseline	4/18/2016
B1110003	Visit 1	4/20/2016
B1110004	Baseline	4/18/2016
B1110004	VISIT 1	4/19/2016
B1110004	VISIT 2	4/21/2016

Need outup;

Output
ID	Vist	Start Date	End Date
B1110001	Visit Group A	4/18/2016	4/19/2016
B1110003	Visit Group A	4/18/2016	4/20/2016
B1110004	Visit Group B	4/18/2016	4/21/2016

Notes;

1.  want to sort same ID and make as one group.

2. assign start date as first visit date.

3. assign end date as last visit date.

sample code;

data want;

set sashelp.class

Please help me.

[PeterClemmensen]

How is Visit Group A / Visit Group B determined?

[tsureshinvites]

attached file shot for your reference,

in that visit of "baseline"  , "c1w1" considered as double-blind treatment. (ie group A)

Visit of "c1w2" considered as open label treatment (ie group B)

so i want to group of (group A and Group B observation as)  epoch as "Treatment".

[Jagadishkatam]

you can try something like below

data have;
input ID$	Vist$	Start_Date$10.;
cards;
B1110001 Baseline 4/18/2016
B1110001 Visit 4/19/2016
B1110003 Baseline 4/18/2016
B1110003 Visit1 4/20/2016
B1110004 Baseline 4/18/2016
B1110004 VISIT1 4/19/2016
B1110004 VISIT2 4/21/2016
;

data want;
length start end $10.;
set have;
retain start;
by id;
if first.id then start=Start_Date;
if last.id then end=Start_Date;
if last.id;
drop Start_Date;
run;

[jvdl · Accepted solution]

Piggybacking on the datalines provided by Jag above, you could also try something like the following, but I am also unsure how you determine whether it's visit group "A" or "B".

data have;
input ID$	Vist$	Start_Date$10.;
cards;
B1110001 Baseline 4/18/2016
B1110001 Visit 4/19/2016
B1110003 Baseline 4/18/2016
B1110003 Visit1 4/20/2016
B1110004 Baseline 4/18/2016
B1110004 VISIT1 4/19/2016
B1110004 VISIT2 4/21/2016
;

proc sql noprint;
	create table want as
		select distinct ID, "Visit Group ?" as visit, min(start_date) as start_date, max(start_date) as end_date
		from have
		group by id
		order by id
	;
quit;

[tsureshinvites]

Thanks a lot. your simple code made wonderful output.

[ballardw]

Please do not post a question in terms different than your "data". Your spreadsheet does not use "Visit Number", so you really need to provide significant detail on how to use the values C1W1 and such that do appear.

It appears that possibly your first step would be to sort the data by subjid date and visit.

Then describe what the desired output should be subjects with no visit other than Baseline. Or if Visits other than Baseline include both C1W1 and C1W2 (or other combinations if possible).

From your example data it appears that you have some subjid that do not have a BASELINE entry, only the C1W2. So what do you expect the output for those to look like?