Longitudinal Data: How do I calculate blood pressure means per subject over multiple visits?

[joesmama]

Hello,

I'm trying to calculate blood pressure means per subject over the course of multiple

blood pressure  readings. Subjects can have a range of visits from few to many. 

My data is organized as follows

PAT_ID	Date	Systolic	Diastolic	Meds
1	2-Nov-22	155	85
1	3-Nov-22	135	90
1	5-Nov-22	145	77
1	6-Nov-22	145	82
2	7-Nov-22	158	98
2	22-Apr-22	130	70
2	22-May-22	140	80	med 1
2	4-Oct-22	130	72
2	11-Oct-22	142	65
3	14-Oct-22	154	61
3	18-Oct-22	145	70
4	23-Oct-22	150	69
4	30-Oct-22	144	71
4	22-Apr-22	165	90	med 2
4	22-Jun-22	160	80	med 2
5	22-Aug-22	.	.	med 3
5	4-Oct-22	128	93
5	8-Oct-22	162	96
5	9-Oct-22	158	96
6	11-Oct-22	149	91
6	14-Oct-22	154	86

I know some of my analysis will involve FIRST and LAST, but how do I calculate the mean per subject when the number of visits per subjects is not necessarily the same. I think I need to use an array here but need a little help getting started. 

Thank you in advance for your help.  

[PaigeMiller · Accepted solution]

---

@joesmama wrote:

I'm trying to calculate blood pressure means per subject over the course of multiple

blood pressure  readings. Subjects can have a range of visits from few to many. 

I know some of my analysis will involve FIRST and LAST, but how do I calculate the mean per subject when the number of visits per subjects is not necessarily the same. I think I need to use an array here but need a little help getting started. 

---

FIRST. and LAST. will work fine in this case, but for cases where you need to calculate statistics for a "group" (in this case a PAT_ID), I would use PROC MEANS

proc means data=have;
    class pat_id;
    var systolic diastolic;
run;

[joesmama]

Thank you Paige. I was definitely over thinking it.