Optimization of antimicrobial dosing in patients with acute kidney injury: a single-centre observational study

Abstract

Abstract Background Acute kidney injury (AKI) is a potential complication of systemic infection. Optimizing antimicrobial dosing in this dynamic state can be challenging with sub- or supra-therapeutic dosing risking treatment failure or toxicity, respectively. Locally, unadjusted renal dosing for the first 48 h of infection is recommended. Objectives To determine the outcomes associated with this dosing strategy. Methods A retrospective cohort analysis was undertaken in patients treated for Gram-negative bacteraemia with concurrent non-filtration dependent AKI from a single-centre NHS acute hospital (April 2016–March 2020). Patient demographics, microbiology data, antimicrobial treatment and patient outcome (in-hospital mortality and kidney function) were analysed. Results In total, 647 episodes of Gram-negative bacteraemia (608 patients) were included; 305/608 (50.2%) were male with median age 71 years (range 18–100). AKI was present in 235/647 (36.3%); 78/647 (12.1%) and 45/647 (7.0%) having Kidney Disease Improving Global Outcomes-defined injury (stage 2) or failure (stage 3), respectively. In-hospital 30 day mortality was 25/352 (7.1%), 14/112 (12.5%), 26/123 (21.1%) and 11/60(18.3%) in patients with normal renal function, AKI stage 1, AKI stage ≥2 and established chronic kidney disease, respectively. Recovery of renal function at Day 21 or discharge was present in 105/106 surviving patients presenting with AKI stage ≥2. Time to recovery of AKI was similar in patients receiving full, low or no aminoglycoside (3 versus 4 versus 3 days, P = 0.612) and those receiving full- and low-dose β-lactam (3 versus 5 days, P = 0.077). Conclusions There is a high burden of AKI in patients with Gram-negative bacteraemia. Dose adjustments of β-lactams may not be necessary in the first 48 h of infection-induced AKI and single-dose aminoglycosides may be considered for early empirical coverage.