Evaluating NDF-RT’s Comprehensiveness as a Drugs Classification Standard: Drug Interaction Checker.

Alotaibi, Awatif and Alkhattabi, Mona (2016) Evaluating NDF-RT’s Comprehensiveness as a Drugs Classification Standard: Drug Interaction Checker. In: Contemporary Consumer Health Informatics. Healthcare Delivery in the Information Age . Springer International Publishing, Australia, pp. 169-198. ISBN 978-3-319-25973-4

[img] PDF
10.1007%2F978-3-319-25973-4_9 - Published Version

Download (54kB)
Official URL: http://link.springer.com/chapter/10.1007%2F978-3-3...


Today, health services are seeing a strong relationship between the number of distributed medications and potential drug–drug interactions for any patient. With this in mind, this project aims to utilize an international drug classification standard and evaluate this standard by implementing a mobile application, “Drug Interaction Checker,” which could be used by patients to check the interactions between their drugs. National Drug File-Reference Terminology application program interface (NDF-RT APIs) were chosen for the appropriate information that they provided to correlate with the information needed by the Drug Interaction Checker. Then, a composition workflow was built between the two NDF-RT API resources to find an appropriate result. An NSXML parser was chosen for the iPhone environment to parse the drug data from XML files by initiating requests using RESTFul Web services. To evaluate the comprehensiveness of NDF-RT, data were collected from 300 experiments from KFMC patients based on two pillars: evaluation of R1 (completeness) and evaluation of R2 (the level of modernization/updating).The data set used to evaluate R1 entailed 116 filtered drugs, 109 retrieved by NDF-RT and 100 retrieved by Lexicomp. The two results were compared. In addition, results were classified into “compatible” and “different” categories: 64 experiments were found to be from the “different” category—there was 1 returned data and 63 unreturned data for NDF-RT and 61 returned data and 3 unreturned data for Lexicomp. The result for R1 was high recall, moderate precision, and moderate accuracy. This is near to the optimal result. The result for R2 was low recall, low precision, and low accuracy, which is very far from the optimal result.

Item Type: Book Section
Subjects: Information Systems
Divisions: Female Community College - Khamis Mushyait > Information Systems
Depositing User: L. Awatif ALOtaibi
Date Deposited: 05 Mar 2017 21:08
Last Modified: 05 Mar 2017 21:08
URI: http://eprints.kku.edu.sa/id/eprint/435

Actions (login required)

View Item View Item