Abstract: SA-PO135
Combining Arterial and Venous Intrarenal Doppler Assessment for the Prediction of AKI After Cardiac Surgery
Session Information
- AKI: Biomarkers, Imaging, Interventions
November 04, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Acute Kidney Injury
- 102 AKI: Clinical, Outcomes, and Trials
Authors
- Giles, Cameron, McMaster University, Hamilton, Ontario, Canada
- Denault, André, Universite de Montreal, Montreal, Quebec, Canada
- Beaubien-Souligny, William, Universite de Montreal, Montreal, Quebec, Canada
Background
Acute kidney injury (AKI) is common after cardiac surgery and often hemodynamically mediated. The roles of ultrasound measures of intrarenal perfusion to predict AKI are yet to be determined. The objective of this study was to determine if point-of-care ultrasound Doppler measures of intrarenal arterial and venous flow predict AKI after cardiac surgery.
Methods
We conducted a secondary analysis of a prospective cohort study of adult patients undergoing cardiac surgery in whom ultrasound assessments were performed at ICU admission after surgery. AKI was defined by the KDIGO creatinine criteria. Intrarenal arterial markers included renal resistive index (RRI) and velocity-time integral normalized to peak systolic velocity (VTI/PSV), while venous markers included intrarenal venous flow (IRVF) categories and renal venous stasis index (RVSI). The area under the receiving operating characteristic (AUROC) curves were used to determine the predictive characteristics for post-operative AKI. The performance of individual markers were compared to a combined RRI and RVSI logistic regression model using the net classification index (NRI) and AUROC were compared with the DeLong test.
Results
We included 131 patients in total, with 47 patients (35.9%) developing post-operative AKI. All studied ultrasound markers showed moderate discrimination for the subsequent development of AKI (Table 1). More complex measurements (VTI/PSV and RVSI) were not superior to simpler indices (RRI and IRVF). In a multivariable model, both RRI (aOR:1.70, CI:1.09-2.66, p=0.02) and RVSI (aOR:0.85, CI:0.04-0.98, p=0.048) remained associated with AKI. The predicted probabilities from the model were slightly better than each index taken individually according to the NRI. However, the AUROC were not significantly different (Table 1).
Conclusion
Intrarenal arterial and venous Doppler indices moderately predict the development of post-operative AKI in cardiac surgery patients. However, combining arterial and venous Doppler indices only marginally improves prediction.
Echographic Parameters and AKI After Cardiac Surgery
| Echographic Parameter | Postoperative Day 0 | DeLong Test/Net Reclassification Index | ||
| AUROC | 95% CI | P Value | ||
| RRI | 0.64 | 0.55-0.74 | 0.006 | p=0.38 / NRI: 0.32 (CI: -0.06; 0.69) |
| VTI/PSV | 0.67 | 0.57-0.77 | 0.001 | |
| IRVF Category | 0.64 | 0.53-0.74 | 0.009 | p=0.19 / NRI: -0.52 (CI: -0.82; 0.20) |
| RVSI | 0.60 | 0.50-0.71 | 0.045 | |
| Predicted probability of model: RRI + RVSI | 0.69 | 0.59-0.78 | <0.001 | Vs RRI: p=0.50 / NRI: 0.50 (CI: 0.17; 0.84) Vs RVSI : p=0.38 / NRI: 0.38 (CI: 0.04; 0.70) |