To calculate the difference of the same variable between adjacent rows of the same patient

[zimcom]

Hi Commuinity,

I have a Vital Signs dataset below which I want to add a column to flag the differences of Systolic Blood Pressure from Baseline to 2 hours, 2 hours to 24 hours. 24 hours to 48 hours and 48 hours to Day 5 for every single patient in the dataset.

Any ideas?

Thank you

Patient	Visit	Systolic Blood Pressure	Diastolic Blood Pressure
1001	Baseline	146	82
1001	hr2	119	59
1001	hr24	133	71
1001	hr48	102	51
1001	D5	128	77
1002	Baseline	122	76
1002	hr2	150	78
1002	hr24	115	61
1002	hr48	121	73
1002	D5	142	85
1003	Baseline	102	59
1003	hr2	110	77
1003	hr24	111	58
1003	hr48	124	78
1003	D5	108	68

[VDD]

proc transform would be the tool I would use.

https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_prinqual_se...

https://stats.idre.ucla.edu/sas/modules/how-to-reshape-data-long-to-wide-using-proc-transpose/

then evaluate the your client across your time progression of  base, 2, 24, 48 and day5 by arraying the time intervals

[zimcom]

@VDD@mkeintz

Thank you both!

[mkeintz · Accepted solution]

You want a sequence of differences between   systolicbp{I}-systolicbp{i-1}, where i is the observations number. And for the first obs of each patient, you want the difference reset to missing.

Use the DIF function, which is defined as    DIF(x)=x-lag(X).

data want;
  set have;
  by patient;
  systolic_diff=dif(systolicbp);
  if first.patient then systolic_diff=.;
run;

Edited addition: I should have added that this program assumes there are no "holes" in the data.  I.e. each patient has a record for every time point, in sequence.