Determinants influencing medication-related alert handling in the electronic health record: A systematic review and meta-analysis.

Abstract

This study aimed to identify and quantify determinants influencing prescribers' handling of medication-related computer-based alerts in electronic health records. A systematic review and meta-analysis were conducted using Embase and Medline databases until January 1^st, 2024. Inclusion criteria were studies in specialist care settings; studying computer-based, medication-related alerts that could be accepted or overridden; and reported a quantitative association between determinants and alert handling. Risk of bias was assessed using the Mixed Methods Appraisal Tool version 2018. Out of 5801 unique articles, 46 were included in the determinant analysis. Most studies were conducted in the United States (63%) and used commercial clinical decision support systems (48%). Five determinant themes emerged: alert-related, medication-related, patient-related, prescriber-related, and setting-related. Seventeen articles were included in the meta-analysis. We found a significant result for increased alert acceptance of fellows compared to residents (OR (95-CI): 1.14 (1.09-1.20)). Furthermore, we found higher odds for acceptance of alerts fired during weekdays when compared to alerts fired during the weekend (OR (95-CI): 1.25 (1.11-1.40)). However, the number of studies combined for both these outcomes was small. We performed a systematic review and meta-analysis of quantitatively reported determinants of influence on alert handling. High risk of bias led to several studies being excluded from the quantitative synthesis. This systematic review suggests a quantifiable relationship between alert handling and several non-modifiable determinants. Our attempt to quantify and meta-analyze these relationships showed strong correlation between a total of eleven individual determinants and alert handling, but due to high heterogeneity consistent effect could only be found in two. These findings can guide further research to improve the evidence quality on determinants influencing alert handling and increase outcome comparability.