Strategies and outcomes of CDSS implementation for antimicrobial stewardship in hospital settings: a systematic review

OBJECTIVES: Optimal implementation strategies and clinical outcomes of Clinical Decision Support Systems (CDSS) for antimicrobial prescribing in hospital settings have not been systemically evaluated. This review explores how CDSS for antimicrobial stewardship (AMS) have been implemented in secondary and tertiary care, focusing on strategies used, clinical and implementation outcomes. METHODS: A systematic search was conducted including studies published up to December 2022. Primary studies describing CDSS implementation strategies in secondary and tertiary care were included. Strategies were analysed using the Expert Recommendations for Implementing Change (ERIC) framework. Implementation outcomes reported in the studies were extracted and categorized according to Proctor's framework. Quality assessment was performed using the Integrated quality Criteria for the Review Of Multiple Study designs (ICROMS). RESULTS: Screening of 2,189 papers identified 12 studies meeting inclusion criteria, all focusing on antimicrobial prescribing in high-income countries. Most CDSS were expert systems (n = 11), primarily designed for infectious disease physicians (n = 7). Pre-implementation assessments, such as workflow analysis, user surveys and multidisciplinary meetings, were conducted in only five of the 12 studies. Studies used a median of 11 out of 73 ERIC implementation strategies. The most frequently reported strategies belonged to the following ERIC categories: developing stakeholder interrelationships (n = 11), training and educating users (n = 11), evaluative strategies (n = 10), provision of interactive assistance (n = 5), adaptation and tailoring to context (n = 5). In contrast, strategies aimed at supporting clinicians (n = 3) and changing infrastructure (n = 2) were less commonly used. No study reported strategies related to patient and service user engagement or financial strategies. Clinical outcomes were considered effective in two studies and partially effective in four, while the remaining studies did not evaluate them. A median of three implementation outcomes was reported per study, with appropriateness (n = 10), adoption (n = 9) and acceptability (n = 9) being the most examined. Overall, initial adoption was slow but improved over time, enhancing compliance with policy indicators. CONCLUSIONS: The implementation strategies of CDSS for AMS in hospital settings are variably reported, with many studies providing limited detail on strategy selection, application, or outcomes, highlighting the need for more systematic and comprehensive evaluation in future research.