Delivered dose of continuous venovenous hemofiltration predicts outcome in septic patients with acute kidney injury: a retrospective study.

PURPOSE In continuous venovenous hemofiltration (CVVH) issues like timing and dose remain controversial, particularly in sepsis. The objective of this study is to examine which CVVH characteristic best predicts mortality in sepsis-induced acute kidney injury (AKI). MATERIALS AND METHODS We retrospectively studied all consecutive patients with sepsis-induced AKI requiring CVVH in a 1.5-year period. Patient, sepsis, and CVVH characteristics, including timing, dose, mode, type of substitution fluid and of anticoagulation, and azotemic control were evaluated. Primary outcome was survival at day 28 after the start of CVVH. RESULTS Of the 97 patients, 43 (44%) died up to day 28 after the start of CVVH. In univariate analyses, the delivered dose of CVVH was about 10% higher in survivors than nonsurvivors (median, 23 vs 20 mL kg(-1) h(-1), P = .01). In multivariate analyses, a lower delivered CVVH dose contributed to predict higher mortality, independently of disease severity, type of substitution fluid, and azotemic control. In a Kaplan-Meier curve, a delivered dose less than 19.7 mL kg(-1) h(-1) was associated with shorter survival (P = .006). CONCLUSION Our retrospective data suggest that in sepsis-induced AKI requiring CVVH, delivered dose, rather than timing, mode of administration, and azotemic control, is an independent predictor of mortality. A lower delivered dose is associated with higher mortality.