A comparative study on the consistency of different decision support software systems from the perspective of potential drug-drug interactions in intensive care unit patients.

Abstract

Background: In intensive care units (ICU), the issue of drug-drug interactions (DDIs) is becoming increasingly prominent, and these interactions can lead to adverse drug reactions, therapeutic failure, or altered drug efficacy. This study aimed to assess the frequency and attributes of potential drug-drug interactions (pDDIs) and the consistency of different decision support software in ICU patients.

Research design and methods: A cross-sectional study was conducted in a tertiary hospital. The consistency of different decision support software was assessed using the Kendall W coefficient, Cohen's kappa, Cronbach's Alpha, Fleiss Kappa, Intraclass correlation coefficient and Gwet's AC1.

Results: A total of 897 prescriptions from 290 patients were evaluated. The total number of pDDIs identified varied significantly across platforms, ranging from 134 to 213. Inter-platform agreement on severity classification was poor (Gwet's AC1 = 0.32, ICC = 0.41).

Conclusions: This observational investigation revealed marked variability across clinical decision platforms regarding both quantitative and qualitative aspects of pDDIs identification in critical care populations, underscoring the imperative to establish unified protocols for pDDIs classification and implement dynamic DDI database maintenance.