Health care and the relational database{1}

by Giselle N
The article I read this week was about relational databases in health care information systems. Of course, medical information is very complex and using it towards diagnosis and treatment decisions requires a lot of intensive data and knowledge. Before the relational database, health care information systems were really just transaction processing systems and had very little decision support capabilities. Now, the systems are able to answer questions by comparing multiple files/tables easily. However, building these databases do not come so easy. It requires a good, in depth data model design in order to know how the connections/relationships are going to work in advance. This is why it is very important for the database designer to know the business very well, for flexibility within the database in the future. The author gives an example of a question a relational database could answer, for example, “How much should the capitation rate–i.e., the per-member-per-month (PMPM) amount–be in the contract?” He goes on to explain that, “Your database has separate tables containing the demographics and as much medical history as possible per member. For some of these members, there are data in other tables documenting their health care–the diagnoses, treatments, lab tests, medications, referrals, and so on. It’s a matter of joining these tables to assemble a total cost with utilization quantities per episode by type of patient, and then comparing these benchmarks to what the health maintenance organization (HMO) or employer’s proposed panel’s statistics look like.” By designing an accurate relational database, one can always know how to make use of these patterns and tables clearly. The author points out that the most important thing and probably the most difficult thing in health care information systems is planning. He says it is important to understand and sometimes even predict. “Health care information systems must be able to respond to questions you may not have thought of the month before.”

I believe the most important thing to take from this article is the importance of creating a data model. The author stresses the need for planning and how important it is to know in advance how relationships are going to work and what not and this can be done accurately through modeling. Learning about the business and understanding it will also make for a good database.


Borok, L. S. (1995). The use of relational databases in health care information systems. Journal of Health Care Finance, 21(4), 6-12. Retrieved from