Artificial intelligence and computerized decision support in adult intensive care: A systematic review of randomized controlled trials

Clinical Snapshot

Key Findings

• P Value: Not reported in abstract

• Effect Size: 8 of 10 RCTs demonstrated improvement in at least one process measure; 2 of 10 RCTs reported reductions in mortality (one sepsis prediction model, one machine-learning-based early-warning system). No pooled effect size reported.

• Primary Outcome: Process-related outcomes (earlier deterioration recognition, protocol adherence, physiological stability) and patient-centred outcomes (mortality, clinical endpoints)

• Nnt Or Sensitivity: Not calculable from available data; no NNT, sensitivity, specificity, or hazard ratios reported at the systematic review level

• Confidence Interval: Not reported in abstract; no meta-analytic pooling performed or reported

Clinical Application

Feasibility of implementation varies substantially by intervention type, institutional infrastructure, and workflow integration. Process improvements are achievable but require sustained clinician engagement and system integration. The evidence does not yet support universal adoption for patient-centred outcome improvement. Pragmatic trials with improved reporting are needed before widespread deployment. In the Australian context, AI and CDSS tools in ICUs are not currently subject to specific PBS listing but may require TGA regulatory approval as Software as a Medical Device (SaMD) under the TGA's Digital Health framework. The Australian Commission on Safety and Quality in Health Care (ACSQHC) and the College of Intensive Care Medicine (CICM) have not yet issued specific guidelines on AI-based CDSS in ICUs. RACGP guidance is not directly applicable to ICU settings. Australian ICUs considering AI/CDSS implementation should note that the evidence base supports process improvements but not consistent patient-centred benefit, and should ensure any deployed system meets TGA SaMD requirements and undergoes local clinical validation before routine use. The Australian Digital Health Agency's national digital health strategy provides a relevant policy framework for integration. Adult patients in general and specialist ICUs where AI-based clinical decision support or CDSS tools are being considered for implementation. Most applicable to high-acuity settings with existing electronic health record infrastructure capable of supporting real-time algorithmic outputs.

Abstract

BACKGROUND: Artificial intelligence (AI) and computerized clinical decision support systems (CDSS) are increasingly applied in intensive care, yet their clinical impact remains uncertain, as most studies focus on model development rather than prospective evaluation. OBJECTIVES: To identify randomized controlled trials (RCTs) evaluating AI-based or CDSS interventions intended to influence real-time decision-making in adult intensive care units (ICUs) and to assess their effects on process and patient-centered outcomes. METHODS: We conducted a systematic review of randomized controlled trials (RCTs). PubMed, Embase, Cochrane CENTRAL, ScienceDirect, and IEEE Xplore were searched from inception to November 2025. Eligible studies evaluated AI-based or CDSS interventions in adult ICU patients. Study quality was assessed using RoB 2 and CONSORT-AI criteria. RESULTS: Ten RCTs were included, enrolling approximately 100,000 adult ICU patients. Five trials evaluated AI-based interventions and five evaluated CDSS. Eight trials showed improvements in at least one process measure, including earlier recognition of deterioration, improved protocol adherence, and physiological stability. However, patient-centered benefits were uncommon. Two trials reported reductions in mortality (a sepsis prediction model and a machine-learning-based early-warning system), while most studies showed no consistent effects on clinical outcomes. Reporting of AI-specific elements-dataset provenance, algorithm versioning, and human-AI interaction-was frequently incomplete. CONCLUSIONS: AI and CDSS interventions in adult ICUs are associated with improvements in process-related outcomes but show limited and inconsistent effects on patient-centered endpoints. These findings highlight a persistent gap between algorithmic innovation and clinical validation and underscore the need for pragmatic randomized trials with improved reporting and integration into clinical workflows.

References

1. 1.Muñoz, J., Fernández-Araujo, N. J., Ruíz-Cacho, R., & Muñoz-Visedo, J. (2026). Artificial intelligence and computerized decision support in adult intensive care: A systematic review of randomized controlled trials. Journal of Critical Care. https://doi.org/10.1016/j.jcrc.2026.155600