Installation:3.2 Indivo Deployment Utilities

Overview
The Indivo Deployment Utilities package includes the Coding System Web Service, all XML-based Indivo coding systems, database setup scripts, and data loading scripts for supported non-XML coding systems.

= Database Setup = A database is required to support non-XML coding systems, i.e. ICD-9-CM and NDC. We provide setup scripts for MySQL and Oracle databases. In the query tool of your choosing, run the setup script that corresponds to your database. The Oracle script will create a user INDIVOCODES -- alter the script before running to specify a password of your choosing. The MySQL script creates a database "indivocodes," and assumes you have or will create a user with access to it.

= Data Loading = With your tables created, you are now ready to load data using the provided Perl scripts. These Perl scripts are fully compatible with both MySQL and Oracle; they only require that you have the Perl modules DBI and either DBD-mysql or DBD-oracle installed. You must also download the code data from the source websites; instructions are included below. Please note that these scripts must be run from a machine with network access to the database you configured above.

Windows Users
Visit ActiveState.com to download ActivePerl, a free, ready-to-install distribution of "Perl, popular modules, the Perl Package Manager (PPM), and complete documentation."

Data Source Names
The loaders require the use of data source names (DSNs) to access source and target databases. Examples follow.

MySQL
dbi:mysql:indivocodes;host=hostname

Oracle
dbi:Oracle:host=hostname;sid=SID

Access via ODBC
dbi:ODBC:datasourcename

NDC data is pulled from an Access database. How to set up an ODBC connection to an Access database is beyond the scope of this document.

ICD-9-CM
Download the ICD9 source data from the CDC PHIN Vocabulary Access website. Click "Express Downloads," then the "Text(tab)" link under "Diagnosis (ICD-9 CM)." Save this to a convenient directory.

The ICD-9 script (icd9.pl) takes four command-line arguments:

$ ./icd9.pl sourcePath targetDsn targetUsername targetPassword


 * sourcePath : Absolute path to the source data text file you downloaded.
 * targetDsn : Data source name of your codes database (see above)
 * targetUsername : Codes database username
 * targetPassword : Codes database password

If the source can be read, and your database is accessible, the script will insert about 13,000 codes -- generally a quick operation.

National Drug Codes
Download the free Cerner Multum Lexicon from multum.com (registration required). Extract the contents of the download and register drug_mlt.mdb as an ODBC data source on your computer.

The NDC script (ndc.pl) takes four command-line arguments:

$ ./ndc.pl sourceDsn targetDsn targetUsername targetPassword


 * sourceDsn : ODBC data source name of your drug database (see above)
 * targetDsn : Data source name of your codes database (see above)
 * targetUsername : Codes database username
 * targetPassword : Codes database password

If the script can properly access your source and target databases, it will proceed to transfer about 150,000 rows. This may take several minutes.