Given on-going needs to aggregate and consolidate data from operational systems and data warehouses, data may be migrated numerous times. How can a health care organization ensure that data quality is maintained and improved using the extract, transform and load (ETL) process? Be sure to support your position with specific examples.

In order for a healthcare organization (HCO) to manage its data as an asset one can employee a data warehouse (DW). Data that is used in a DW must be extracted from operational systems. Operational data houses disparate data from multiple source systems that must be integrated prior to loading into DW, e.g. clinical, financial, registration, on-line transaction processing (OLTP), etc. (Anonymous, 2000). Since you shouldn’t directly work with operational data, a working copy of the data will be needed for manipulation without impacting other systems. Extraction, transform, load (ETL) systems will extract from operational systems and create a fixed-in-time snap shot of the data (Miron, 2011).

ETL is one of the most challenging and risky steps in quality data management but one that should never be overlooked. The goal of the data extraction process is to bring all source data into a common, consistent format so it can be made ready for loading into the data warehouse. This stage is so crucial to the DW as this is where most of the data is cleaned, as different source systems can have variation in format, different source codes for the same kind of data, invalid characters, etc., it is these issues that make ETL necessary to transform data into useable, consistent and reliable form for loading into the DW (Kakish & Kraft, 2012).

Healthcare organizations (HCO) should invest in IT resources that can succeed at this level of work, if this resource is overlooked there may be a “dump and load” pile of data, views, tables,