The confounding role of common diabetes medications in developing acute renal failure: A data mining approach with emphasis on drug-drug interactions

Behrooz Davazdahemami, Dursun Delen

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

Longstanding diabetes mellitus is today known as the primary reason for kidney failure in the patients having that condition. While the prior research has studied the confounding role of some frequently prescribed diabetes medications in developing acute renal failure, some rarely prescribed medications are still under-studied in this regard. In addition, even for those drugs studied in the past, inconsistent findings have been reported. In the present study, by extending a data mining framework from the prior research and equipping that with some standard statistical metric from the medical literature we investigate the general confounding role of the common diabetes medications in developing acute renal failure in a large group of patients with diabetes mellitus (Type II). In addition, we assess the stability of the identified confounding roles by taking into account the potential drug-drug interactions between those diabetes medications with a group of drugs already known to have negative effect on the kidney function. Our results suggest the general dominant confounding role for each of the diabetes medications, but also suggests that these roles are unstable across various prescription combinations due to potential drug-drug interactions, thereby provide an explanation for the inconsistent findings in the literature.

Original languageEnglish
Pages (from-to)168-177
Number of pages10
JournalExpert Systems with Applications
Volume123
DOIs
StatePublished - 1 Jun 2019
Externally publishedYes

Keywords

  • Acute renal failure
  • Adverse drug reactions
  • Diabetes
  • Drug-drug interactions
  • Itemset mining

Fingerprint

Dive into the research topics of 'The confounding role of common diabetes medications in developing acute renal failure: A data mining approach with emphasis on drug-drug interactions'. Together they form a unique fingerprint.

Cite this