Dear Epidata Team, We have a patient's database. A patient can have multiplediagnosis. Hence these are the numerically coded variables defined for dataentry of ;Diag, Diag2, Diag3, Diag4, Diag5, Diag6. Of course one can FREQ theIndividual "diag" and see the individual diagnosis pattern. However if onewants to see the pattern of common diagnostic combinations observed inall the patients, is there any way out? Dr. Shavinder Singh
You can create new, temporary files with the values for each ‘diag’ as a new variable like this: * EpiData Analysis V2.2 syntax! * adiag will hold all diagnoses eventually define adiag #### * get all first diagnoses adiag = diag * save all records with non-missing diag1 select adiag <> . savedata diag1.rec /replace select * save all records with non-missing diag2 adiag = diag2 select adiag <> . savedata diag2.rec /replace … savedata diag6.rec /replace close * read / append all temporary files read "diag1.rec” append /file="diag2.rec” … append /file="diag6.rec” * get overall frequencies freq adiag
Jamie
On Nov 25, 2017, at 07:18, EpiData development and support epidata-list@lists.umanitoba.ca wrote:
Dear Epidata Team, We have a patients database. A patient can have multiple diagnosis. Hence these are the numerically coded variables defined for dataentry of Diag, Diag2, Diag3, Diag4, Diag5, Diag6. Of course one can FREQ the Individual 'diag'; and see the individual diagnosis pattern. However if one wants to see the pattern of common diagnostic combinations observed in all the patients, is there any way out? Dr. Shavinder Singh _______________________________________________ EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list
To see the combinations you may supplement Jamie's suggestion with this
Define alldiag _________.... Make 80 underscores
Alldiag = diag1 + diag2 + etc.
Freq alldiag
Best wishes Jens Lauritsen
Den 25. november 2017 16.16.33 CET, EpiData development and support epidata-list@lists.umanitoba.ca skrev:
You can create new, temporary files with the values for each ‘diag’ as a new variable like this:
- EpiData Analysis V2.2 syntax!
- adiag will hold all diagnoses eventually
define adiag ####
- get all first diagnoses
adiag = diag
- save all records with non-missing diag1
select adiag <> . savedata diag1.rec /replace select
- save all records with non-missing diag2
adiag = diag2 select adiag <> . savedata diag2.rec /replace … savedata diag6.rec /replace close
- read / append all temporary files
read "diag1.rec” append /file="diag2.rec” … append /file="diag6.rec”
- get overall frequencies
freq adiag
Jamie
On Nov 25, 2017, at 07:18, EpiData development and support
epidata-list@lists.umanitoba.ca wrote:
Dear Epidata Team, We have a patients database. A patient can have multiple diagnosis. Hence these are the numerically coded variables defined for dataentry
of
Diag, Diag2, Diag3, Diag4, Diag5, Diag6. Of course one can FREQ the Individual 'diag'; and see the individual diagnosis pattern. However if one wants to see the pattern of common diagnostic
combinations observed in
all the patients, is there any way out? Dr. Shavinder Singh _______________________________________________ EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list
EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list
In my response, I did not consider that you would be looking for patterns of diagnoses that occurred together.
You might add this to my previous program
freq adiag // as before tables adiag diag1 tables adiag diag2 …
This will result in large tables, but will allow you to look at cells where there seem to be higher frequencies.
Jamie
On Nov 25, 2017, at 10:26, EpiData development and support epidata-list@lists.umanitoba.ca wrote:
To see the combinations you may supplement Jamie's suggestion with this
Define alldiag _________.... Make 80 underscores
Alldiag = diag1 + diag2 + etc.
Freq alldiag
Best wishes Jens Lauritsen
participants (1)
-
EpiData development and support