1. Improved diagnosis-medication association mining to reduce pseudo-associations.
- Author
-
Wang, Ching-Huan, Nguyen, Phung Anh, (Jack) Li, Yu Chuan, Islam, Md. Mohaimenul, Poly, Tahmina Nasrin, Tran, Quoc-Viet, Huang, Chih-Wei, and Yang, Hsuan-Chia
- Subjects
- *
ASSOCIATION rule mining , *ALGORITHMS , *DATA science , *SERVER farms (Computer network management) - Abstract
Association rule mining has been adopted to medical fields to discover prescribing patterns or relationships among diseases and/or medications; however, it has generated unreasonable associations among these entities. This study aims to identify the real-world profile of disease-medication (DM) associations using the modified mining algorithm and assess its performance in reducing DM pseudo-associations. We retrieved data from outpatient records between January 2011 and December 2015 in claims databases maintained by the Health and Welfare Data Science Center, Ministry of Health and Welfare, Taiwan. The association rule mining's lift (Q-value) was adopted to quantify DM associations, referred to as Q 1 for the original algorithm and as Q 2 for the modified algorithm. One thousand DM pairs with positive Q 1 -values (Q 1 +) and negative or no Q 2 -values (Q 2 − or Q 2 ∅) were selected as the validation dataset, in which two pharmacists assessed the DM associations. A total of 3,120,449 unique DM pairs were identified, of which there were 333,347 Q 1 + Q 2 − pairs and 429,931 Q 1 + Q 2 ∅ pairs. Q 1 + Q 2 − rates were relatively high in ATC classes C (29.91%) and R (30.24%). Classes L (69.91%) and V (52.52%) demonstrated remarkably high Q 1 + Q 2 ∅ rates. For the 1000 pairs in the validation, 93.7% of the Q 1 + Q 2 − or Q 1 + Q 2 ∅ DM pairs were assessed as pseudo-associations. However, classes M (5.3%), H (4.5%), and B (4.1%) showed the highest rates of plausible associations falsely given Q 2 − or Q 2 ∅ by the modified algorithm. The modified algorithm demonstrated high accuracy to identify pseudo-associations regarded as positive associations by the original algorithm and would potentially be applied to improve secondary databases to facilitate research on real-world prescribing patterns and further enhance drug safety. • Association rule mining's lift was used to measure disease-medication associations. • Over one billion prescriptions retrieved from claims databases were analyzed. • Our modified algorithm showed high accuracy to reduce pseudo-associations. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF