Many healthcare organizations find themselves coming across
the following problems caused by data quality: fraud, bad debt, billing
inefficiencies, and even life threatening issues that involve diagnosis and
prescribing care and real-time analytics.
On top of all that, healthcare organizations are attempting
to handle the increase in the amounts of data while transitioning from paper to
electronic health records.
This leads healthcare organizations to search for a data
warehousing solution that integrates data quality into all of their
Here are four recommended steps to eliminate the root causes
of data problems before they affect your data warehousing and BI efforts.
Step 1: Profile Your Data
It is easy to undermine the affects of data quality issues
revolving around deduplication, incompleteness, and inconsistencies. Profiling
your data occurs at the start and analyzes the quality of data values within
and across data sets.
Step 2: Verify, Correct, Update and Standardize Your Data
It is crucial to identify data issues and determine the best
approach to cleaning and updating your data. With contact data always being in
flux, outdated data has become a major issue. This results in undelivered mail
as well as costs associated with returned mail and wasted postage.
Another big issue with a huge impact is fraud. Fraud costs
the healthcare industry nearly $70 billion annually according to the National
Health Care Anti-Fraud Association. This problem can be solved by verifying
patient data at the point of care and maintaining it updated throughout the
Step 3: Fill in the Gaps of Missing Information
Adding missing email addresses phone numbers, and completing
addresses increase patient engagement, lowers costs, improves quality, and
retains patients and members.
Step 4: Merge, Eliminate and Consolidate Duplicate Records
Duplicate records undermine the integrity of a data
warehouse while also proving to be very costly. Duplicate pair of records
creates $50 in hidden operational costs resulting in $78,000 per year,
according to the American Medical Informatics Association. Not to mention the
harm in duplicates. A provider can easily mistake one patient for another with
a similar name.