Machine Learning-Guided Fluid Resuscitation for Acute Pancreatitis Improves Outcomes

Niwen Kong; Patrick Chang; Shulman Ira; Ubayd Haq; Maziar Amini; Denis Nguyen; Farhaad Khan; Rachan Narala; Nisha Sharma; Daniel Wang; Tiana Thompson; Jonathan Sadik; Cameron Breze; David C Whitcomb; James L Buxbaum

doi:10.14309/ctg.0000000000000825

Machine Learning-Guided Fluid Resuscitation for Acute Pancreatitis Improves Outcomes

Clin Transl Gastroenterol. 2025 Jan 24. doi: 10.14309/ctg.0000000000000825. Online ahead of print.

Authors

Affiliations

¹ Division of Gastroenterology, Department of Medicine.
² Department of Pathology, University of Southern California.
³ Ariel Precision Medicine.
⁴ Division of Gastroenterology, Department of Medicine, University of Pittsburgh.

PMID: 39851257
DOI: 10.14309/ctg.0000000000000825

Abstract

Objectives: Ariel Dynamic Acute Pancreatitis Tracker (ADAPT) is an artificial intelligence tool using mathematical algorithms to predict severity and manage fluid resuscitation needs based on the physiologic parameters of individual patients. Our aim was to assess whether adherence to ADAPT fluid recommendations versus standard management impacted clinical outcomes in a large prospective cohort.

Method: We analyzed patients consecutively admitted to the Los Angeles General Medical Center between June 2015 to November 2022 whose course was richly characterized by capturing more than 100 clinical variables. We inputted this data into the ADAPT system to generate resuscitation fluid recommendations. and compared to the actual fluid resuscitation within the first 24 hours from presentation. The primary outcome was the difference in organ failure in those who were over (>500cc)- versus adequately (within 500cc) resuscitated with respect to the ADAPT fluid recommendation. Additional outcomes included ICU admission, SIRS at 48 hours, local complications, and pancreatitis severity.

Results: Among the 1083 patients evaluated using ADAPT, 700 were over-resuscitated,196 were adequately resuscitated, and 187 were under-resuscitated. Adjusting for pancreatitis etiology, gender, and SIRS at admission, over-resuscitation was associated with increased respiratory failure (Odd Ratio (OR) 2.73 [95%CI 1.06, 7.03]) as well as ICU admission (OR 2.40 [1.41, 4.11]), more than 48 hours of hospital length of stay (OR 1.87 [1.19, 2.94]), SIRS at 48 hours (OR 1.73 [1.08, 2.77]) and local pancreatitis complications (OR 2.93 [1.23, 6.96]).

Conclusions: Adherence to ADAPT fluid recommendations reduces respiratory failure and other adverse outcomes compared to conventional fluid resuscitation strategies for acute pancreatitis. This validation study demonstrates the potential role of dynamic machine learning tools in acute pancreatitis management.