Kidney Week

Abstract: FR-PO023

Loop Diuretic Challenge to Predict the Need for Renal Replacement Therapy Among Patients with Stage III AKI

Session Information

• AKI: Clinical, Outcomes, Trials - I
  October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Acute Kidney Injury

• 102 AKI: Clinical, Outcomes, and Trials

Authors

• Sakhuja, Ankit, Mayo Clinic, Rochester, Minnesota, United States

Ankit Sakhuja,

• Barreto, Erin F., Mayo Clinic, Rochester, Minnesota, United States

Erin F. Barreto,

• Albright, Robert C., Mayo Clinic, Rochester, Minnesota, United States

Robert C. Albright,

• Kashani, Kianoush, Mayo Clinic, Rochester, Minnesota, United States

Kianoush Kashani,

Background

A poor response to a high dose loop diuretic challenge (LDC) predicts the progression of stage I and II acute kidney injury (AKI). Performance of this diagnostic test in stage III AKI is unknown. The objective of this study was to determine if poor response to LDC among patients with stage III AKI predicted the need for renal replacement therapy within 24-hours (dialysis_24h).

Methods

We included adults (≥18 years) admitted or transferred to medical or general surgical ICUs at Mayo Clinic, Rochester, MN between Jan 1, 2004, to Dec 31, 2016. Patients with stage III AKI were identified by an electronic surveillance tool using AKIN criteria. We then identified patients who received LDC, defined as at least 1mg/kg intravenous bolus dose of furosemide or equivalent intravenous bolus dose of bumetanide after the diagnosis of stage III AKI. We excluded patients with the end-stage renal disease, organ transplantation or who died within 24 hours of LDC. We modeled post-LDC urine output as a restricted cubic spline and compared the area under the curve (AUC) for urine output (mL) at 2h (UOP_2h) and 6h (UOP_6h) after LDC to predict dialysis_24h.

Results

We included 687 patients with stage III AKI who received LDC. The patients who received dialysis_24h were younger (63.9 ± 14.5 years vs. 67.6 ± 15.6 years, p= 0.008), had lower Charlson comorbidity indices (4.9±2.4 vs. 5.8±2.7, p< 0.001) and higher SOFA scores on the day of LDC (10.4±3.9 vs. 8.2±3.4, p< 0.001). Both median total 2 hours (UOP_2h) and 6 hours (UOP_6h) urine output after LDC were lower in patients who needed dialysis_24h (UOP_2h 48ml vs. 138ml, p<.001; UOP_6h 210ml vs. 616ml, p< 0.001) but UOP_6h was better in predicting dialysis_24h (Area under the curve 0.71 vs. 0.67, p= 0.02). The sensitivity and specificity of a UOP_6h cutoff of ≤600ml to predict dialysis_24h was 80.9% & 50.5% and for a cutoff of ≤300cc was 64.2% & 68.2%.

Conclusion

Among patients with AKI stage III the UOP_6h after loop diuretic challenge had a modest discriminant capacity to identify dialysis initiation within the next 24 hours. Though its predictive power seems better in earlier stages of AKI, LDC may be a useful adjunct to assess dialysis needs in patients with advanced AKI.

Funding

• Other NIH Support