Comparative effectiveness of albumin versus no albumin on renal replacement therapy and mortality in patients with septic shock and renal impairment.

BACKGROUND Albumin infusions may be renally protective or harmful in patients with septic shock who have kidney impairment. This can affect the need for renal replacement therapy (RRT) and in-hospital mortality. RESEARCH QUESTION Does the early use of albumin mitigate the need for RRT or in-hospital mortality in patients with septic shock and kidney impairment on hospital admission. STUDY DESIGN AND METHODS This was a retrospective, multicenter, inverse probability-of-treatment weighted cohort study conducted in 220 geographically diverse community and teaching hospitals across the U.S. Adult patients were included if they had septic shock and kidney impairment on hospital admission. Patients were categorized as those who received albumin (within 24h of admission) or no albumin during hospitalization. Proportion of patients with RRT or in-hospital mortality were compared between groups. RESULTS Of the 9988 patients included in the final cohort, 7929 did not receive albumin and 2059 received albumin. Patients had a mean (SD) age of 67.8 years (14.8), 46.3% were female, and mean (SD) eGFR was 32 (12) ml/min/1.73m2 on the day of admission. In the weighted cohort, the composite outcome of RRT or in-hospital mortality occurred in 33.8% without albumin and 39.7% with albumin (OR 1.29, 95% CI 1.14 - 1.47, p<0.001). There was no significant difference with 5% albumin (OR 1.07, 95% CI 0.84 - 1.37), but there was a significantly increased risk with 25% albumin (OR 1.43, 95% CI 1.16 - 1.76). INTERPRETATION In patients with septic shock and kidney impairment on hospital admission, early albumin use may be associated with an increased composite outcome of RRT or in-hospital mortality. This increased risk is most associated with hyperoncotic rather than iso-oncotic albumin.