ASN's Mission

To create a world without kidney diseases, the ASN Alliance for Kidney Health elevates care by educating and informing, driving breakthroughs and innovation, and advocating for policies that create transformative changes in kidney medicine throughout the world.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on X

Kidney Week

Please note that you are viewing an archived section from 2023 and some content may be unavailable. To unlock all content for 2023, please visit the archives.

Abstract: SA-OR06

Machine Learning-Guided Personalized Diuretic Strategy in Patients with Sepsis-Associated AKI

Session Information

Category: Acute Kidney Injury

  • 101 AKI: Epidemiology, Risk Factors, and Prevention

Authors

  • Oh, Wonsuk, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Kittrell, Hannah, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Kohli-Seth, Roopa D., Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Nadkarni, Girish N., Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Sakhuja, Ankit, West Virginia University, Morgantown, West Virginia, United States
Background

Fluid overload is common in patients with sepsis associated acute kidney injury (SA-AKI) and is associated with worse outcomes. Loop diuretics are commonly used to manage fluid overload, but their role in patients with SA-AKI is unclear. In this study we aimed to develop a personalized strategy for use of loop diuretics in patients with SA-AKI using machine learning.

Methods

This was a retrospective study using MIMIC IV database. We defined AKI using both creatinine and urine output based KDIGO criteria. We identified patients with sepsis who developed AKI within 48 hours of ICU admission. The primary outcome was AKI recovery within 48 hours of AKI onset. We used available features for demographics, comorbidities, SOFA score, vital signs, laboratory measurements, fluid balance, vasopressors and mechanical ventilation to estimate time varying individual treatment effects (ITE) using a two-model approach employing gradient boosting. The policy tree algorithm with pruning was then employed to identify subpopulations with the highest average treatment effects (ATE) for loop diuretic therapy, enabling a personalized diuresis strategy.

Results

Of 10,739 patients with SA-AKI, 37.8% had AKI recovery within 48 hours. Loop diuretics were used in 3,661 patients within 48 hours after onset of SA-AKI. ATE of loop diuretics was HR 0.935 (95% CI: 0.934, 0.937). Further policy tree analysis identified a personalized strategy for use of loop diuretics with goal to increase AKI recovery within 48 hours (Fig 1a). Specifically, patients with SA-AKI with Blood Urea Nitrogen (BUN)<=18 and red cell distribution width (RDW)<=14.2 benefited from loop diuretic therapy with an ATE of HR 1.086 (CI: 1.080, 1.091) (Fig 1b).

Conclusion

In this study we identified subgroups of patients with SA- AKI who may benefit from loop diuretic therapy to improve the likelihood of recovery of AKI within next 48 hours. This study shows the potential of machine learning to help personalize therapies for patients with AKI.

Funding

  • NIDDK Support