Kidney Week

Abstract: FR-OR101

Personalized Recommendations for AKI Using a Kidney Action Team: A Multicenter Randomized Controlled Trial

Session Information

• High-Impact Clinical Trials - 1
  October 25, 2024 | Location: Hall D, Convention Center
  Abstract Time: 11:00 AM - 11:15 AM

Category: Acute Kidney Injury

• 102 AKI: Clinical, Outcomes, and Trials

Authors

• Aklilu, Abinet Mathias, Yale University School of Medicine, New Haven, Connecticut, United States

Abinet Mathias Aklilu, MD, MPH

• Menez, Steven, Johns Hopkins University, Baltimore, Maryland, United States

Steven Menez, MD

• Yamamoto, Yu, Yale University School of Medicine, New Haven, Connecticut, United States

Yu Yamamoto, MS

• Bitzel, Jack, Johns Hopkins University, Baltimore, Maryland, United States

Jack Bitzel

• Corona Villalobos, Celia Pamela, Johns Hopkins University, Baltimore, Maryland, United States

Celia Pamela Corona Villalobos, MD, MS

• O'Connor, Kyle D., Yale University School of Medicine, New Haven, Connecticut, United States

Kyle D. O'Connor

• Shelton, Kyra A., Yale University School of Medicine, New Haven, Connecticut, United States

Kyra A. Shelton

• Faulkner, Sophia, Yale University School of Medicine, New Haven, Connecticut, United States

Sophia Faulkner

• Makhijani, Amrita, Yale University School of Medicine, New Haven, Connecticut, United States

Amrita Makhijani

• Kadhim, Bashar A., Yale University School of Medicine, New Haven, Connecticut, United States

Bashar A. Kadhim, MD, MHS

• Parikh, Chirag R., Johns Hopkins University, Baltimore, Maryland, United States

Chirag R. Parikh, MD, PhD, FASN

• Wilson, Francis Perry, Yale University School of Medicine, New Haven, Connecticut, United States

Francis Perry Wilson, MD, MS

Background

AKI alerts have not improved patient outcomes but AKI care bundles have. We investigated if timely personalized recommendations for AKI are more effective at improving patient outcomes.

Methods

KAT-AKI was a 1:1 randomized investigator-blinded trial conducted at 2 large US hospital systems, Yale and Johns Hopkins, evaluating the effectiveness of rapid, expert-led diagnosis and management recommendations at creatinine-based AKI diagnosis. A physician and pharmacist “Kidney Action Team”(KAT) screened patients with new AKI at the time of an automated AKI alert that was not visible to clinicians. For patients who passed screening, the KAT remotely reviewed their chart and provided structured recommendations in 5 major categories (general diagnostics, volume, potassium, acidosis, medications). Patients were then randomized to either receive a KAT Recommendation Note in their chart or no KAT Note. The primary clinical outcome was a composite of AKI progression, dialysis or mortality within 14 days. The primary process outcome was proportion of recommendations implemented within 24h of randomization.

Results

The trial included 4003 patients 786(20%) of whom were in an ICU. The median(IQR) age was 72(61-81)yrs, 47% were female and 41% had CKD. The median(IQR) time from automated AKI detection to randomization was 56(33-83)min. Nearly all (96%) had at least one general recommendation and 55% had at least one medication-specific recommendation. The process outcome was significantly higher in the intervention arm (0.25[0-0.50] vs 0[0-0.33], p<0.001). The intervention particularly improved ordering of urinalysis, orthostatic vitals, creatine kinase, obstruction evaluation, AKI documentation and medication changes.(Figure) However, there was no difference in the primary clinical outcome, individual components or post-AKI hospital LOS.(Table)

Conclusion

Personalized early recommendations improved AKI documentation and process outcomes but did not change clinical outcomes.(NCT04040296)

Funding

• Other U.S. Government Support