High speed clinical data retrieval system with event time sequence feature: with 10 years of clinical data of Hamamatsu University Hospital CPOE

From Clinfowiki
Revision as of 04:58, 30 November 2015 by URumana (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

This is a review of an article by Kimura et al, Creating High speed clinical data retrieval system with event time sequence feature: with 10 years of clinical data of Hamamatsu University Hospital CPOE [1]

Objective

To implement a high speed data retrieval system for hospital information accumulated in 10 years for the purpose of research, evidence creation and patient safety.

Methods

Lab results, prescription and injection orders, disease diagnoses, and patients' demographics are exported from Hospital Information System in HL7 format. A hierarchical variable length data structure using MUMPS is used to store data. A total of 73, 709, 298 records is available for searching.

Results

The data retrieval system D*D provides fast retrieval to end user. Simple retrieval takes about 2 seconds. Complex retrievals take around 2 minutes allowing Boolean operations. Average monthly access of the D*D system averages 703 searches. The whole system does not affect routine operation of existing CPOE system.

Conclusion

Using exported data from CPOE, the new system is proved to provide excellent search environment with high speed.

Comments

This paper introduced the D*D system for fast retrieval of CPOE system. The strategy that data was first exported from CPOE and then independently accessed from a cache database has its benefits of not interrupting routine CPOE usage. However, it may not always get the most current data since all data indexed is staged, not real time.


References

  1. Kimura, M., Tani, S., Watanabe, H., Naito, Y., Sakusabe, T., Nakaya, J., ... & Furuta, T. (2008). High speed clinical data retrieval system with event time sequence feature. Methods of information in medicine, 47(6), 560-568. http://www.ncbi.nlm.nih.gov/pubmed/19023490.