categorical

[Anthony Santella]

Dear Listserv,

I would like your thoughts on an issue I am having with a dataset.

Column A lists the admitting diagnoses of all people in my dataset by it's
ICD-9 code (numerical code that is unique for every disease).

Column B: I have a separate list of 20 specific diseases and their ICD-9
codes. I would like to create a new variable in this column that tells me
if any of the diagnoses in Column A are found in this list of 20 diseases.

Any help would be much appreciated.

Thanks in advance!

[Art Kendall]

see if something like one of the snippets of untested syntax is what you
need.
If not please ask a more detailed question.

compute haveanyof20= any(columna, 'a111111111', 'b2222222', 'ccca11111'
. . .'zzzaaa123').



numeric heart lungs gastro, nervous, . .  . psychiatric(n1)
do repeat code = 'a111111111', 'b2222222', 'ccca11111' . . .'zzzaaa123' /
  diagnosis = heart to psychiatric.
compute diagnosis = columna eq code .
end repeat.

Art Kendall
Social Research Consultants


Anthony Santella wrote:

>Dear Listserv,
>
>I would like your thoughts on an issue I am having with a dataset.
>
>Column A lists the admitting diagnoses of all people in my dataset by it's
>ICD-9 code (numerical code that is unique for every disease).
>
>Column B: I have a separate list of 20 specific diseases and their ICD-9
>codes. I would like to create a new variable in this column that tells me
>if any of the diagnoses in Column A are found in this list of 20 diseases.
>
>Any help would be much appreciated.
>
>Thanks in advance!
>
>
>
>

[Maguin, Eugene]

Anthony,

You say:
Column A lists the admitting diagnoses of all people in my dataset by it's
ICD-9 code (numerical code that is unique for every disease).

Column B: I have a separate list of 20 specific diseases and their ICD-9
codes. I would like to create a new variable in this column that tells me
if any of the diagnoses in Column A are found in this list of 20 diseases.

Does this mean that each person has 20 records with the first having the
admitting dx and the first of the 20 specific disease codes and each of the
remaining 19 records having missing data for the admit dx and one of the
remaining 19 disease codes? So if your sample had 100 persons in it, your
dataset would have 2000 records. Could this be true?

Or, do you mean that columns B1 thru B20 carry the 20 preset disease codes?
Thus one record per case rather than 20.

Gene Maguin

[Marks, Jim]

COMPUTE colB = ANY(colA,'788.2','E111').

You will need to supply the list of ICD-9 codes into the any statement.


--jim

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Anthony Santella
Sent: Saturday, November 11, 2006 1:22 PM
To: [hidden email]
Subject: categorical

Dear Listserv,

I would like your thoughts on an issue I am having with a dataset.

Column A lists the admitting diagnoses of all people in my dataset by
it's
ICD-9 code (numerical code that is unique for every disease).

Column B: I have a separate list of 20 specific diseases and their ICD-9
codes. I would like to create a new variable in this column that tells
me if any of the diagnoses in Column A are found in this list of 20
diseases.

Any help would be much appreciated.

Thanks in advance!