***ICD Code Extraction.
*This syntax (mostly) automates the writing of a long syntax program based on information from an Excel file (code and label).
*Version: 10-27-08.

*Improvements, suggestions, etc. to David.Laflamme@unh.edu.

*Import the Excel file first -- you need two columns (variables).
*Vars="code" and "label".

*Strip decimals.
compute code=replace(code,".", "").
exe.

*Create syntax to count each ICD9 code in principle diagnosis field and 9 secondary diagnosis fields.
string syntax (a500).
compute syntax=concat("count var",rtrim(code),"=principal_diagnosis_cde secondary_diag1_cde",
 " secondary_diag2_cde secondary_diag3_cde secondary_diag4_cde secondary_diag5_cde",
 " secondary_diag6_cde secondary_diag7_cde secondary_diag8_cde secondary_diag9_cde ('",rtrim(code),"').").
exe.

*Create variable label syntax.
string labels (a500).
compute labels=concat("variable labels var",rtrim(code)," '",rtrim(code),":",rtrim(label),"'.").

*Output variable names for other programming (i.e. make available a list for cutting and pasting).
string varnames (a500).
compute varnames=concat("var",rtrim(code)).
exe.

*Output variable names with plus sign for other programming (i.e. make available a list for cutting and pasting).
string varnames2 (a500).
compute varnames2=concat("var",rtrim(code)," +").
exe.

*Manually cut and paste the string values of the new variables, starting with the new variable called "syntax", into a syntax file.
*Run the new program on the hospital discharge dataset.