Biomarker- Or Phenotype-Guided Sedation in The Intensive Care Unit: A Systematic Review and Meta-Analysis of Patient-Centered Outcomes

Introduction: Current practices of sedation in intensive care units (ICUs) are inconsistent, often leading to over- or under-sedation, which are associated with adverse outcomes. Personalized sedation seeks to optimize regimens by incorporating individual characteristics, including pharmacogenomics, electroencephalogram (EEG) patterns, and disease phenotypes. This systematic review and meta-analysis summarizes recent evidence on phenotype- and biomarkerguided sedation in the ICU. Methods: Eligible studies included randomized controlled trials (RCTs) and observational studies evaluating biomarkeror phenotype-guided sedation with standard care in adult ICU patients. Searches were conducted in PubMed/MEDLINE, Embase, Cochrane CENTRAL, and ClinicalTrials.gov through May 1, 2025. Mean differences (MDs), standardized mean differences (MDs) or risk ratios (RRs) with 95% confidence intervals (CIs) were pooled using random-effects models in RevMan . The I2 statistic was used to calculate heterogeneity. Results: Five studies met inclusion criteria. Protocolized and biomarker-guided strategies reduced sedative and opioid use, particularly for propofol, midazolam, and morphine. Structured monitoring and analgesia-first approaches were associated with reduced delirium, but no difference for ventilation duration or ICU length of stay. Mortality did not differ significantly, and no increase in adverse events was observed. Conclusion: Biomarker- and phenotype-guided sedation reduces sedative exposure and delirium risk, while remaining safe and without negative impact on mortality.