Calculating mean of multiple patient visits per specialty/year

[Treeva]

Hello SAS Community members,

I am working on patient data for a work project, & the excerpt below shows some of these data:

PatientID	Year_of_Visit	Specialty
1111	2005	Specialist 2
1111	2005	Specialist 2
1111	2005	Specialist 3
1111	2005	Specialist 3
1111	2006	Specialist 2
1111	2006	Specialist 2
1112	2006	Specialist 2
1112	2006	Specialist 2
1112	2007	Specialist 2
1112	2007	Specialist 2
1112	2007	Specialist 6
1112	2007	Specialist 6
1112	2009	Specialist 2
1112	2009	Specialist 2
1112	2009	Specialist 2
1112	2009	Specialist 5
1112	2009	Specialist 6

As you can note, the same patient may see many specialists in the same year.

I am required to calculate the following:

1. Mean of the visits per year

2. Mean of the visits by the specialty type per year

Honesty, I have no idea how to do those calculations due to my limited experience in SAS! I looked into related topics but my question was not fitting any!

I'd be greatly appreciating your kind inputs!

Thanks in advance!

[ballardw · Accepted solution]

Do you want the mean of the count of visits or mean of something else.

Proc summary data=have;

   class specialty year;

   output out=work.counts;

run;

will create several counts with the variable _type_ indicating which group it belongs to.

_type_= 0 will have the overall count, _type_ = 1 and 2 will have summaries by specialty and year only and _type_=3 will have combinations of specialty and year.

Use that output data set to take means of the _freq_ variable by _type_ using any of Proc summary, means, report or tabulate.

[Treeva]

Thanks so much, ballardw!

Your suggestion successfully answered my first question.

For my second question, I had to run proc means by type & specialty to get the mean of visits per year for each specialty.

This is how I used your code:

Proc summary data=freqVisits;

class Specialty Year_of_Visit;

output out=work.counts;

proc print; run;

_type_ = 2 gives results similar to proc freq; tables Specialty;

_type_ = 3 stratifies the frequency of visits to each specialty by year.

The following to show how the proc summary statement worked:

Specialty	Year_of_Visit	_TYPE_	_FREQ_
Specialist 2	.	2	77
Specialist 7	.	2	1
Specialist 3	.	2	17
Specialist 2	2003	3	1
Specialist 2	2004	3	1
Specialist 2	2007	3	5
Specialist 2	2008	3	3
Specialist 2	2009	3	5
Specialist 2	2010	3	11
Specialist 2	2011	3	11
Specialist 2	2012	3	10
Specialist 2	2013	3	16
Specialist 2	2016	3	11
Specialist 2	2017	3	3
Specialist 7	2010	3	1
Specialist 3	2010	3	2
Specialist 3	2011	3	1
Specialist 3	2013	3	1
Specialist 3	2014	3	2
Specialist 3	2015	3	11

/* To calculate (mean ± SD) of visits per year */

proc means data = counts;

var _freq_ ;

where _type_ = 3;

run; 

The result was the following:

Analysis Variable: _FREQ_
N	Mean	Std Dev	Minimum	Maximum
17	5.5882352	4.9378490	1.0000000	16.0000000

/* To calculate (mean ± SD)  of visits per year by specialty type*/

proc means data = counts;

var _freq_ ;

by Specialty ;

where _type_ = 3;

run;

The results were the following:

Specialty=Specialist 2

Analysis Variable: _FREQ_
N	Mean	Std Dev	Minimum	Maximum
11	7.0000000	5.0000000	1.0000000	16.0000000

Specialty = Specialist 7

Analysis Variable: _FREQ_
N	Mean	Std Dev	Minimum	Maximum
1	1.0000000	.	1.0000000	1.0000000

Specialty = Specialist 3

Analysis Variable: _FREQ_
N	Mean	Std Dev	Minimum	Maximum
5	3.4000000	4.2778499	1.0000000	11.0000000

Thanks so much again!