Development and Validation of a Policy Tree Approach for Optimizing Intravenous Fluids in Critically Ill Patients with Sepsis and Acute Kidney Injury

ImportanceIntravenous (IV) fluids are cornerstone for management of acute kidney injury (AKI) after sepsis but can cause fluid overload. Restrictive fluid strategy may benefit some patients, however, identifying them is challenging. Novel causal machine learning (ML) techniques can estimate heterogenous treatments effects (HTE) of IV fluids among these patients. ObjectivesTo develop and validate causal ML framework to identify patients who benefit from restrictive fluids (<500mL fluids within 24 hours after AKI). Design, Setting and ParticipantsWe conducted a retrospective study among patients with sepsis who developed AKI within 48 hours of ICU admission. We developed a causal ML approach to estimate individualized treatment effects and guide fluid therapy. We developed the model in MIMIC-IV and externally validated in SICdb. Main Outcomes and MeasuresOur primary outcome was early AKI reversal at 24 hours. Secondary outcomes included sustained AKI reversal and major adverse kidney events by 30 days (MAKE30). Model performance to identify HTE of restrictive IV fluids was assessed using area under targeting operator characteristic curve (AUTOC), that quantifies how well a model captures HTE and compared to random forest model. ResultsCausal forest model outperformed random forest in identifying HTE of restrictive IV fluids with AUTOC 0.15 vs. -0.02 in external validation cohort. Among 1,931 patients in external validation cohort, the model recommended restrictive fluids for 68.9%. Among these, patients who received restrictive fluids demonstrated significantly higher rate of early AKI reversal (53.9% vs 33.2%, p<0.001), sustained AKI reversal (34.2% vs 18.0%, p<0.001) and lower rates of MAKE30 (17.1% vs 34.6%, p=0.003). Results were consistent in adjusted analysis. Conclusions and RelevanceCausal ML framework outperformed random forest model in identifying patients with AKI and sepsis who benefit from restrictive fluid therapy. This provides a data-driven approach for personalized fluid management and merits prospective evaluation in clinical trials. KEY POINT.O_ST_ABSQuestionC_ST_ABSCan causal machine learning (ML) identify which critically ill, septic patients with AKI benefit from a restrictive fluid management strategy? FindingsIn this retrospective cohort study, we developed and externally validated a causal-ML model to identify septic patients with AKI who would benefit from restrictive IV fluid therapy. Those who received the model-recommended restrictive IV fluids experienced significantly higher rates of early AKI reversal (53.9 % vs. 33.2 %; p<0.001) and lower rates of MAKE 30 (17.1% vs 34.6%, p=0.003). MeaningThis study demonstrates that causal ML can effectively identify septic patients with AKI who are most likely to benefit from a restrictive IV fluids strategy.