Find population frequency when multiple response with long strings

* QUESTION:
I am trying to calculate frequency of responses for the same value for
multiple variables using 'string' category. Specifically, using ICD-9 codes
as the primary cause for hospitalization in a population of 200 patients
that has multiple hospitalizations and multiple reasons for hospitalization.

Example: patientA has been hospitalized 8 times: 2x 'heart attack', 1x
'kidney stone', 4x 'dehydration', 1x 'stroke'. PatientB : 1x 'pneumonia', 1x
'stroke', 1x 'stomach ulcer'. PatientC: 1x 'dehyration', 1x 'heart
attack'..etc

How do I create frequency of response for the entire population?

* SOLUTION posted to SPSSX-L list by rlevesque@videotron.ca on 2001/06/04.


DATA LIST LIST /id(A8) cause1(A12) cause2(A12) cause3(A12) cause4(A12).
BEGIN DATA
'patientA'  'heart attack''heart attack' 'kidney stone''dehydration'
'patientB' 'pneumonia' 'stroke' 'stomach ulcer' 
'patientC'  'dehyration' 'heart attack'
END DATA.
LIST.

* Reformat data.
VECTOR cause=cause1 TO cause4.
STRING cause(A12).
LOOP cnt=1 TO 4.
COMPUTE cause=cause(cnt).
XSAVE OUTFILE='c:\\temp\\temp.sav' /KEEP=id cause.
END LOOP.
EXECUTE.

GET FILE='c:\\temp\\temp.sav'.
SELECT IF LENGTH(RTRIM(cause))>0.
AUTORECODE cause /INTO cause1.
FREQUENCIES
  VARIABLES=cause1
  /ORDER=  ANALYSIS .