RESUMO
BACKGROUND: Several factors influence the outcomes in acute kidney injury (AKI), especially in intensive care unit (ICU) patients. In this scenario, continuous renal replacement therapies (CRRT) are used to control metabolic derangements and blood volume. Knowing this fact, it may be possible to change the course of the disease and decrease the high mortality rate observed. Thus, we aimed to evaluate the main risk factors for death in AKI patients needing CRRT. RESULTS: This was a prospective, observational cohort study of ICU patients (N = 183) with AKI who underwent continuous venovenous hemodiafiltration (CVVHDF) as their initial dialysis modality choice. The patients were predominantly male (62.8%) and their median age was 65 (55-76) years. The most frequent comorbidities were cardiovascular disease (39.3%), hypertension (32.8%), diabetes (24%), and cirrhosis (20.7%). The main cause of AKI was sepsis (52.5%). At beginning of CVVHDF, 152 patients (83%) were using vasopressors. The median SAPS 3 and SOFA score at ICU admission was 61 (50-74) and 10 (7-12), respectively. The dialysis dose delivered was 33.2 (28.9-38.7) ml/kg/h. The median time between ICU admission and CVVHDF initiation was 2 (1-4) days. The median cumulative fluid balance during the CVVHDF period was -1838 (-5735 +2993) ml. The mortality rate up to90 days was 58%. The independent mortality risk factors in propensity score model were: chronic obstructive pulmonary disease (OR = 3.44[1.14-10.4; p = 0.028]), hematologic malignancy (OR = 5.14[1.66-15.95; p = 0.005]), oliguria (OR = 2.36[1.15-4.9; p = 0.02]), positive daily fluid balance during CVVHDF (OR = 4.55[2.75-13.1; p<0.001]), and total SOFA score on first dialysis day (OR = 1.27[1.12-1.45; p<0.001]). CONCLUSIONS: Dialysis-related factors may influence the outcomes. In our cohort, positive daily fluid balance during CRRT was associated with lower survival. Multicenter, randomized studies are needed to assess fluid balance as a primary outcome to define the best strategy in this patient population.
