Artificial Intelligence–Enhanced Tele-Nursing for Early Clinical Deterioration Detection in Critical Care: A Systematic Review and Meta-Analysis

Introduction: Early identification of clinical deterioration in intensive care units (ICUs) remains a global challenge, directly influencing morbidity, mortality, and resource utilization. Despite advances in monitoring technology, the timely recognition of subtle physiological changes often depends on human interpretation. Integrating artificial intelligence (AI) algorithms with tele-nursing systems may bridge this gap by enabling continuous, data-driven surveillance and decision support. This study systematically reviews and meta-analyzes current evidence on the effectiveness of AI-enhanced tele-nursing interventions in detecting early deterioration among critically ill patients.Methods: Following PRISMA ۲۰۲۰ guidelines, five electronic databases (PubMed, Scopus, Web of Science, CINAHL, and Cochrane Library) were searched for studies published between ۲۰۱۰ and ۲۰۲۵. Eligible articles included randomized controlled trials (RCTs) and observational studies comparing AI-integrated tele-nursing systems with conventional monitoring in ICU settings. Data were independently screened and extracted by two reviewers. Methodological quality was appraised using the Cochrane Risk of Bias tool (RoB ۲.۰) for RCTs and the Newcastle-Ottawa Scale (NOS) for observational studies. A random-effects meta-analysis (DerSimonian-Laird method) pooled standardized mean differences (SMD) for detection accuracy and risk ratios (RR) for ICU mortality. Heterogeneity was assessed using the I^۲ statistic.Results: Preliminary synthesis of ۱۸ eligible studies (۵ RCTs, ۱۳ observational; ۵,۲۰۰ patients) demonstrated that AI-driven tele-nursing significantly improved early-warning detection accuracy (SMD = ۰.۶۸; ۹۵% CI ۰.۴۴–۰.۹۱; p < ۰.۰۰۱). Furthermore, the intervention group showed a statistically significant reduction in ICU mortality (RR = ۰.۸۵; ۹۵% CI ۰.۷۷–۰.۹۴; p = ۰.۰۰۲), representing a ۱۵% relative risk reduction compared with routine care. Heterogeneity was moderate for accuracy (I^۲ = ۴۲%) and low for mortality (I^۲ = ۲۱%). Sensitivity analyses confirmed result robustness.Conclusion: AI-enhanced tele-nursing represents a transformative approach to critical-care monitoring, enabling earlier intervention and improving patient outcomes. Findings highlight the need for large-scale clinical integration and nurse-training frameworks to ensure safe, ethical, and sustainable deployment of intelligent monitoring systems. This review underscores the pivotal role of nurses in shaping technology-driven critical-care innovation.

کلیدواژه ها:

Artificial Intelligence ، Tele-nursing ، Critical Care Nursing ، Clinical Deterioration ، Meta-Analysis