EACTS Congenital Database

Since 2002, I'm involved in EACTS Congenital Database. At first, I've migrated the central database from MySQL to PostgreSQL and re-designed it. For example, in old version of the database structure, the date of death was bound to operation instead of patient. So one patient who had many operations could have several different dates of death.

During migration to PostgreSQL, I had to resolve many inconsistency problems, like: references to non-existing patients, references to non-existing operations, missing data and many similar. I tried not to throw out any data; to mend as much as possible during migration.

I've implemented the data validation. It employs a XML-based configuration for easy validation rules' addition and removal.

Another thing was a web interface for data entry. I can't provide a direct link because it's available only for project participants.

Since 2005, a database change was introduced. I redesigned the database once more, taking into account two years of experience with the database, focusing on data integrity and possible future analysis. A new data collection software was written. Software was written in Java by one of my coworkers, following my guidelines. Again, I designed the database for the software. Software was released under GNU General Public License. It uses embedded Java database, Derby.

I created a Wiki website with project documentation and downloads. At first it was based on MoinMoin, but then I migrated it to MediaWiki, enhanced navigation and added more diagrams.

EACTS Database is the longest and the largest project that I participated in.


Author: automatthias

You won't believe what a skeptic I am.