JMIR Publications

ABSTRACT

Background:

Clinical decision support system (CDSS) has been appeared as an indispensable tool for reducing medication errors and adverse drug events. A significant amount of studies reported that medication-related alerts in the current CDSS were often inappropriately overridden. Increased overridden rate has already raised questions about the appropriateness of CDSS application and raised a concern about the patient’s safety and quality of care.

Objective:

We, therefore, aim to conduct a systematic review to determine the overridden rate, the reasons for the alerts override and evaluate the appropriateness of overrides reasons.

Methods:

Systematic articles search of Google Scholar, PubMed, Embase, Scopus, Web of Science without language restrictions between January 1, 2000, and March 31, 2019. Two authors independently extracted data and cross-checked to avoid errors. Finally, the quality of the included studies was examined following the Cochrane guidelines.

Results:

A total of 23 articles were included in our systematic review. The rage of average override alerts was 46.2% to 96.2%. However, the rage of the overall appropriateness rate for override was 29.4% ~ 100%. Although, the rage of appropriateness was varied by alert types (Drug-allergy: 63.4%~ 100%, drug-drug interaction: 0%~95%, dose: 43.9%~88.8%, geriatric:14.3%~57%, renal: 27%~87.5%. Indeed, the interrater reliability for the assessment of override alerts appropriateness was excellent (the range of κ was 0.79~0.97). The most common reasons for override were “will monitor”, “patients have tolerated before”.

Conclusions:

The findings of our study show that CDSS overridden rate was often high and most of the cases overrides were identified as appropriate but the rate of appropriateness varied widely by different types of alerts. However, inappropriate overrides were associated with a higher rate of potential ADEs when compared with appropriate overrides. Future efforts should try to focus on reducing alert fatigue by providing alerts that are clinically relevant, information related alerts are clear and unambiguous.