{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,5,2]],"date-time":"2026-05-02T08:23:12Z","timestamp":1777710192877,"version":"3.51.4"},"posted":{"date-parts":[[2024,8,7]]},"group-title":"Intensive Care and Critical Care Medicine","reference-count":48,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2024,8,7]],"date-time":"2024-08-07T00:00:00Z","timestamp":1722988800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.medrxiv.org\/about\/FAQ#license"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,10,13]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                <jats:sec>\n                  <jats:title>Importance<\/jats:title>\n                  <jats:p>Intravenous (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.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Objectives<\/jats:title>\n                  <jats:p>To develop and validate causal ML framework to identify patients who benefit from restrictive fluids (&lt;500mL fluids within 24 hours after AKI).<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Design, Setting and Participants<\/jats:title>\n                  <jats:p>We 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.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Main Outcomes and Measures<\/jats:title>\n                  <jats:p>Our 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.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>Causal 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&lt;0.001), sustained AKI reversal (34.2% vs 18.0%, p&lt;0.001) and lower rates of MAKE30 (17.1% vs 34.6%, p=0.003). Results were consistent in adjusted analysis.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusions and Relevance<\/jats:title>\n                  <jats:p>Causal 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.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>KEY POINT.<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Question<\/jats:title>\n                    <jats:p>Can causal machine learning (ML) identify which critically ill, septic patients with AKI benefit from a restrictive fluid management strategy?<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Findings<\/jats:title>\n                    <jats:p>In 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&lt;0.001) and lower rates of MAKE 30 (17.1% vs 34.6%, p=0.003).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Meaning<\/jats:title>\n                    <jats:p>This study demonstrates that causal ML can effectively identify septic patients with AKI who are most likely to benefit from a restrictive IV fluids strategy.<\/jats:p>\n                  <\/jats:sec>\n                <\/jats:sec>","DOI":"10.1101\/2024.08.06.24311556","type":"posted-content","created":{"date-parts":[[2024,8,7]],"date-time":"2024-08-07T17:10:51Z","timestamp":1723050651000},"source":"Crossref","is-referenced-by-count":5,"title":["Personalized Fluid Management in Patients with Sepsis and AKI: A Casual Machine Learning Approach"],"prefix":"10.64898","author":[{"given":"Wonsuk","family":"Oh","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kullaya","family":"Takkavatakarn","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zainab","family":"Al-taie","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hannah","family":"Kittrell","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Khaled","family":"Shawwa","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hernando","family":"Gomez","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1525-8541","authenticated-orcid":false,"given":"Ashwin S.","family":"Sawant","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Pranai","family":"Tandon","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Gagan","family":"Kumar","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Michael","family":"Sterling","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ira","family":"Hofer","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Lili","family":"Chan","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"John","family":"Oropello","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Roopa","family":"Kohli-Seth","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8135-6858","authenticated-orcid":false,"given":"Alexander W","family":"Charney","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Monica","family":"Kraft","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Patricia","family":"Kovatch","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mayte","family":"Su\u00e1rez-Fari\u00f1as","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"John A.","family":"Kellum","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Girish N.","family":"Nadkarni","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ankit","family":"Sakhuja","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2025101508050720000_2024.08.06.24311556v3.1","doi-asserted-by":"publisher","DOI":"10.1007\/s00134-008-1367-2"},{"key":"2025101508050720000_2024.08.06.24311556v3.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.kint.2019.05.026"},{"key":"2025101508050720000_2024.08.06.24311556v3.3","doi-asserted-by":"publisher","DOI":"10.1007\/s00134-015-3850-x"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.4","doi-asserted-by":"crossref","first-page":"165","DOI":"10.1186\/s13054-016-1343-0","article-title":"Urine sodium concentration to predict fluid responsiveness in oliguric ICU patients: a prospective multicenter observational study","volume":"20","year":"2016","journal-title":"Crit Care"},{"issue":"2","key":"2025101508050720000_2024.08.06.24311556v3.5","doi-asserted-by":"crossref","first-page":"e0192592","DOI":"10.1371\/journal.pone.0192592","article-title":"Aggressive fluid accumulation is associated with acute kidney injury and mortality in a cohort of patients with severe pneumonia caused by influenza A H1N1 virus","volume":"13","year":"2018","journal-title":"PLoS One"},{"issue":"5","key":"2025101508050720000_2024.08.06.24311556v3.6","doi-asserted-by":"crossref","first-page":"431","DOI":"10.1097\/SHK.0000000000000453","article-title":"Increased Fluid Administration After Early Acute Kidney Injury is Associated with Less Renal Recovery","volume":"44","year":"2015","journal-title":"Shock"},{"issue":"6","key":"2025101508050720000_2024.08.06.24311556v3.7","doi-asserted-by":"crossref","first-page":"780","DOI":"10.1111\/aas.13105","article-title":"Fluid accumulation during acute kidney injury in the intensive care unit","volume":"62","year":"2018","journal-title":"Acta Anaesthesiol Scand"},{"key":"2025101508050720000_2024.08.06.24311556v3.8","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMoa062200"},{"key":"2025101508050720000_2024.08.06.24311556v3.9","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMoa2202707"},{"key":"2025101508050720000_2024.08.06.24311556v3.10","doi-asserted-by":"publisher","DOI":"10.1007\/s00134-021-06401-6"},{"key":"2025101508050720000_2024.08.06.24311556v3.11","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-02902-1"},{"issue":"4","key":"2025101508050720000_2024.08.06.24311556v3.12","doi-asserted-by":"crossref","first-page":"710","DOI":"10.1002\/cpt.3159","article-title":"Using Machine Learning to Individualize Treatment Effect Estimation: Challenges and Opportunities","volume":"115","year":"2024","journal-title":"Clin Pharmacol Ther"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.13","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/s41597-022-01899-x","article-title":"MIMIC-IV, a freely accessible electronic health record dataset","volume":"10","year":"2023","journal-title":"Sci Data"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.14","doi-asserted-by":"crossref","first-page":"320","DOI":"10.1038\/s41597-024-03164-9","article-title":"Harnessing Big Data in Critical Care: Exploring a new European Dataset","volume":"11","year":"2024","journal-title":"Sci Data"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.15","doi-asserted-by":"crossref","first-page":"156","DOI":"10.1186\/s13054-024-04935-x","article-title":"Machine learning derived serum creatinine trajectories in acute kidney injury in critically ill patients with sepsis","volume":"28","year":"2024","journal-title":"Crit Care"},{"key":"2025101508050720000_2024.08.06.24311556v3.16","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2016.0287"},{"key":"2025101508050720000_2024.08.06.24311556v3.17","doi-asserted-by":"publisher","DOI":"10.1038\/kisup.2012.1"},{"key":"2025101508050720000_2024.08.06.24311556v3.18","doi-asserted-by":"publisher","DOI":"10.1038\/nrneph.2017.2"},{"key":"2025101508050720000_2024.08.06.24311556v3.19","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMoa1711584"},{"issue":"22","key":"2025101508050720000_2024.08.06.24311556v3.20","doi-asserted-by":"crossref","first-page":"1869","DOI":"10.1093\/ajhp\/zxz203","article-title":"Evidence-based development of a nephrotoxic medication list to screen for acute kidney injury risk in hospitalized children","volume":"76","year":"2019","journal-title":"Am J Health Syst Pharm"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.21","doi-asserted-by":"crossref","first-page":"212","DOI":"10.1186\/s13054-025-05457-w","article-title":"ORAKLE: Optimal Risk prediction for mAke30 in patients with sepsis associated AKI using deep LEarning","volume":"29","year":"2025","journal-title":"Crit Care"},{"issue":"3","key":"2025101508050720000_2024.08.06.24311556v3.22","first-page":"1","article-title":"mice: Multivariate Imputation by Chained Equations inR","volume":"45","year":"2011","journal-title":"Journal of Statistical Software"},{"issue":"523","key":"2025101508050720000_2024.08.06.24311556v3.23","doi-asserted-by":"crossref","first-page":"1228","DOI":"10.1080\/01621459.2017.1319839","article-title":"Estimation and Inference of Heterogeneous Treatment Effects using Random Forests","volume":"113","year":"2018","journal-title":"Journal of the American Statistical Association"},{"key":"2025101508050720000_2024.08.06.24311556v3.24","doi-asserted-by":"publisher","DOI":"10.1093\/aje\/kwq439"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.25","doi-asserted-by":"crossref","first-page":"133","DOI":"10.3982\/ECTA15732","article-title":"Policy Learning With Observational Data","volume":"89","year":"2021","journal-title":"Econometrica"},{"issue":"549","key":"2025101508050720000_2024.08.06.24311556v3.26","doi-asserted-by":"crossref","first-page":"38","DOI":"10.1080\/01621459.2024.2393466","article-title":"Evaluating Treatment Prioritization Rules via Rank-Weighted Average Treatment Effects","volume":"120","year":"2025","journal-title":"J Am Stat Assoc"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.27","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random Forests","volume":"45","year":"2001","journal-title":"Machine Learning"},{"key":"2025101508050720000_2024.08.06.24311556v3.28","doi-asserted-by":"publisher","DOI":"10.1136\/bmj-2023-078378"},{"key":"2025101508050720000_2024.08.06.24311556v3.29","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMra032401"},{"issue":"6","key":"2025101508050720000_2024.08.06.24311556v3.30","doi-asserted-by":"crossref","first-page":"924","DOI":"10.1177\/0310057X0703500611","article-title":"May CN: A comparison of 4% succinylated gelatin solution versus normal saline in stable normovolaemic sheep: global haemodynamic, regional blood flow and oxygen delivery effects","volume":"35","year":"2007","journal-title":"Anaesth Intensive Care"},{"key":"2025101508050720000_2024.08.06.24311556v3.31","doi-asserted-by":"publisher","DOI":"10.1097\/ALN.0b013e3181c4a5e2"},{"key":"2025101508050720000_2024.08.06.24311556v3.32","doi-asserted-by":"publisher","DOI":"10.1378\/chest.124.3.1053"},{"key":"2025101508050720000_2024.08.06.24311556v3.33","doi-asserted-by":"publisher","DOI":"10.1164\/rccm.201511-2285OC"},{"key":"2025101508050720000_2024.08.06.24311556v3.34","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajpath.2011.10.011"},{"key":"2025101508050720000_2024.08.06.24311556v3.35","doi-asserted-by":"publisher","DOI":"10.1152\/ajprenal.00687.2010"},{"key":"2025101508050720000_2024.08.06.24311556v3.36","doi-asserted-by":"publisher","DOI":"10.1186\/2110-5820-1-27"},{"key":"2025101508050720000_2024.08.06.24311556v3.37","doi-asserted-by":"publisher","DOI":"10.1038\/nrneph.2016.186"},{"key":"2025101508050720000_2024.08.06.24311556v3.38","doi-asserted-by":"publisher","DOI":"10.1016\/S0140-6736(04)16815-3"},{"issue":"4","key":"2025101508050720000_2024.08.06.24311556v3.39","first-page":"860 e867","article-title":"Associations of fluid overload with mortality and kidney recovery in patients with acute kidney injury: A systematic review and meta-analysis","volume":"30","year":"2015","journal-title":"J Crit Care"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.40","doi-asserted-by":"crossref","first-page":"17410","DOI":"10.1038\/s41598-023-44778-0","article-title":"Associations between fluid overload and outcomes in critically ill patients with acute kidney injury: a retrospective observational study","volume":"13","year":"2023","journal-title":"Sci Rep"},{"issue":"Suppl 3","key":"2025101508050720000_2024.08.06.24311556v3.41","first-page":"S94","article-title":": Fluid Overload and Acute Kidney Injury","volume":"24","year":"2020","journal-title":"Indian J Crit Care Med"},{"key":"2025101508050720000_2024.08.06.24311556v3.42","doi-asserted-by":"crossref","first-page":"668688","DOI":"10.3389\/fvets.2021.668688","article-title":"Fluid Overload","volume":"8","year":"2021","journal-title":"Front Vet Sci"},{"issue":"1","key":"2025101508050720000_2024.08.06.24311556v3.43","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1186\/s13054-019-2534-2","article-title":"Intravenous fluid resuscitation is associated with septic endothelial glycocalyx degradation","volume":"23","year":"2019","journal-title":"Crit Care"},{"key":"2025101508050720000_2024.08.06.24311556v3.44","doi-asserted-by":"crossref","first-page":"100094","DOI":"10.1016\/j.mbplus.2021.100094","article-title":"Endothelial glycocalyx degradation during sepsis: Causes and consequences","volume":"12","year":"2021","journal-title":"Matrix Biol Plus"},{"issue":"3","key":"2025101508050720000_2024.08.06.24311556v3.45","doi-asserted-by":"crossref","first-page":"149","DOI":"10.4103\/IJCIIS.IJCIIS_17_18","article-title":"Moradimajd P: The relationship between fluid resuscitation and intra-abdominal hypertension in patients with blunt abdominal trauma","volume":"8","year":"2018","journal-title":"Int J Crit Illn Inj Sci"},{"issue":"4","key":"2025101508050720000_2024.08.06.24311556v3.46","doi-asserted-by":"crossref","first-page":"e32807","DOI":"10.1097\/MD.0000000000032807","article-title":"Intra-abdominal hypertension among medical septic patients associated with worsening kidney outcomes (IAH-WK study)","volume":"102","year":"2023","journal-title":"Medicine (Baltimore"},{"issue":"9","key":"2025101508050720000_2024.08.06.24311556v3.47","first-page":"e45620","article-title":"The Effects of Restrictive Fluid Resuscitation on the Clinical Outcomes in Patients with Sepsis or Septic Shock: A Meta-Analysis of Randomized-Controlled Trials","volume":"15","year":"2023","journal-title":"Cureus"},{"issue":"10","key":"2025101508050720000_2024.08.06.24311556v3.48","doi-asserted-by":"crossref","first-page":"1172","DOI":"10.1111\/acem.14546","article-title":"Restrictive fluids versus standard care in adults with sepsis in the emergency department (REFACED): A multicenter, randomized feasibility trial","volume":"29","year":"2022","journal-title":"Acad Emerg Med"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2024.08.06.24311556","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T20:08:46Z","timestamp":1768507726000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2024.08.06.24311556"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,8,7]]},"references-count":48,"URL":"https:\/\/doi.org\/10.1101\/2024.08.06.24311556","relation":{},"subject":[],"published":{"date-parts":[[2024,8,7]]},"subtype":"preprint"}}