Fluid responsiveness in ARDS: current evidence, knowledge gaps, and future directions — a scoping review of the literature

BACKGROUND: Predicting fluid responsiveness in patients with acute respiratory distress syndrome (ARDS) remains a clinical challenge. While dynamic indices such as pulse pressure variation (PPV), stroke volume variation (SVV), and functional tests like passive leg raising (PLR), end-expiratory occlusion test (EEOT), and tidal volume challenge (VTC) are widely used in critical care, their reliability in ARDS-characterized by altered lung mechanics and heterogeneous pathophysiology-has not been clearly established. OBJECTIVE: To systematically map the available evidence on predictors of fluid responsiveness in ARDS, including their performance under different ventilatory settings, the use of extracorporeal support, and the influence of respiratory mechanics and phenotypic variability. METHODS: A scoping review was conducted following PRISMA-ScR and Arksey & O'Malley guidelines. A systematic search of PubMed, Embase, Web of Science, and Cochrane Library was performed from January 1999 to October 2024. Studies were included if they evaluated any fluid responsiveness predictor in adult ARDS patients and reported performance metrics such as sensitivity, specificity, or area under the curve (AUC). Data were extracted in duplicate and synthesized narratively. RESULTS: Ten studies were included. PPV was the most frequently evaluated predictor, followed by SVV, EEOT, VTC, and PLR. Only two studies assessed patients in the prone position, and two included patients supported with veno-venous extracorporeal membrane oxygenation (VV-ECMO). No study included spontaneously breathing patients or stratified results by ARDS phenotype. Diagnostic performance varied substantially across clinical contexts. Unadjusted PPV showed inconsistent diagnostic performance, particularly in patients with low respiratory system compliance or under prone positioning, whereas adjusted PPV, defined as PPV normalized to respiratory mechanics (e.g., driving pressure, transpulmonary pressure, respiratory system compliance, and chest wall elastance), demonstrated improved diagnostic performance in selected cohorts. Overall, predictor performance was highly dependent on ventilatory strategy, extracorporeal support, and validation methodology. CONCLUSION: The diagnostic performance of fluid responsiveness predictors in ARDS is highly variable and context-dependent, influenced by ventilatory strategy, respiratory mechanics, extracorporeal support, and methodological heterogeneity. Important evidence gaps persist, particularly in prone ventilation, VV-ECMO, and spontaneously breathing patients. Future studies should integrate measures of fluid tolerance and stratify patients by ARDS phenotype to support individualized fluid management strategies.