Association between intrarenal venous flow from Doppler ultrasonography and acute kidney injury in patients with sepsis in critical care: a prospective, exploratory observational study.

Background: Intrarenal venous flow (IRVF) patterns assessed using Doppler renal ultrasonography are real-time bedside visualizations of renal vein hemodynamics. Although this technique has the potential to detect renal congestion during sepsis resuscitation, there have been few studies on this method. We aimed to examine the relationship between IRVF patterns, clinical parameters, and outcomes in critically ill adult patients with sepsis. We hypothesized that discontinuous IRVF was associated with elevated central venous pressure (CVP) and subsequent acute kidney injury (AKI) or death.

Methods: We conducted a prospective observational study in two tertiary-care hospitals, enrolling adult patients with sepsis who stayed in the intensive care unit for at least 24 h, had central venous catheters placed, and received invasive mechanical ventilation. Renal ultrasonography was performed at a single time point at the bedside after sepsis resuscitation, and IRVF patterns (discontinuous vs. continuous) were confirmed by a blinded assessor. The primary outcome was CVP obtained at the time of renal ultrasonography. We also repeatedly assessed a composite of Kidney Disease Improving Global Outcomes of Stage 3 AKI or death over the course of a week as a secondary outcome. The association of IRVF patterns with CVP was examined using Student's t-test (primary analysis) and that with composite outcomes was assessed using a generalized estimating equation analysis, to account for intra-individual correlations. A sample size of 32 was set in order to detect a 5-mmHg difference in CVP between IRVF patterns.

Results: Of the 38 patients who met the eligibility criteria, 22 (57.9%) showed discontinuous IRVF patterns that suggested blunted renal venous flow. IRVF patterns were not associated with CVP (discontinuous flow group: mean 9.24 cm HO [standard deviation: 3.19], continuous flow group: 10.65 cm HO [standard deviation: 2.53], p = 0.154). By contrast, the composite outcome incidence was significantly higher in the discontinuous IRVF pattern group (odds ratio: 9.67; 95% confidence interval: 2.13-44.03, p = 0.003).

Conclusions: IRVF patterns were not associated with CVP but were associated with subsequent AKI in critically ill adult patients with sepsis. IRVF may be useful for capturing renal congestion at the bedside that is related to clinical patient outcomes.