Databases Reference
In-Depth Information
FIGURE B.7
Common traits of data warehouse.
the decryption key to determine the real identification of the patient. In such a manner there may be
feedback from the collective common data warehouse to the provider that does not violate the privacy
of the individual patient.
Common elements of a data warehouse
There are several recognizable architectural elements to the collective common data warehouse
(indeed, these recognizable elements are for
any
data warehouse.) The first of those elements is that
data is represented in a single integrated manner in the data warehouse. The second element is that
data can be collected over time in the collective common data warehouse. Whereas the hospital OLTP
and administrative systems contain reasonably current information, the data warehouse contains his-
torical information, say five to ten years' worth of information. The third element of a data warehouse
is that data is at a low level of granularity. These are the three distinguishing characteristics of data in
a data warehouse. Depending on the particulars a data warehouse may or may not contain identify-
ing information, depending on the compliance specifications.
Figure B.7
shows the general, common
characteristics of a data warehouse.
One of the issues that is important as data is sent to the data warehouse is that of the protection
of privacy. As a rule, individual identifying data is safe when stored at the transaction level. But as
data is placed into a data warehouse, care must be taken not to violate the rules laid down by Health
Insurance Portability and Accountability Act (HIPAA). It is noted that the research done on the data
found in the data warehouse is normally not dependent on individual identification. When an analyst
looks at 10,000 cases of a disease and its symptoms, the analyst is normally not interested in knowing
who the individuals are who have contributed information to the study. Instead, the analyst is much
more interested in the characteristics, the treatment, the symptoms, and the outcomes of the disease.
Therefore, placing data in the data warehouse anonymously is an acceptable practice.
