Impact of Guideline-Based Acute Kidney Injury Care Processes on Outcomes in Hospitalized Patients With Decompensated Cirrhosis

GOALS: We evaluated whether adherence to acute kidney injury (AKI) specific process measures (PMs) improves outcomes in hospitalized patients with cirrhosis and AKI. BACKGROUND: Expert bodies recommend PMs for the management of AKI in cirrhosis, but the link between PMs and outcomes is unclear. STUDY: We performed a retrospective cohort study of patients hospitalized with ascites and AKI at Veterans Administration hospitals. We examined the relationship between PMs [avoidance of diuretics, nonselective beta-blockers (NSBB), and other vasodilators, volume expansion with albumin/crystalloids, infectious work-up, subspecialty consultation, and postdischarge outpatient visit] and patient outcomes. Outcomes were in-hospital AKI resolution and 90-day mortality, adjusted for patient demographics, comorbidities, cirrhosis complications, Model for End-Stage Liver Disease (MELD) 3.0 score, and AKI stage. RESULTS: Among 2221 patients, adherence to PMs varied: NSBB, diuretics, and other vasodilators were avoided in 97%, 79%, and 80% of cases, respectively. Volume expansion with albumin and crystalloids was administered in 21% and 29% of cases, while 47% had an infectious workup and 29% had subspecialty consultation. Postdischarge follow-up rates within 15 days were low: 28% for primary care, 6% for gastroenterology/hepatology, and 1% for nephrology. Overall, 53% achieved AKI resolution, and 40% died within 90 days. Diuretic avoidance was associated with increased likelihood of AKI resolution (OR: 1.14, 95% CI: 1.02-1.28). Timely postdischarge outpatient visits with primary care [hazard ratio (HR): 0.53, 95% CI: 0.43-0.67] or gastroenterology/hepatology (HR: 0.44, 95% CI: 0.29-0.66) were independently associated with reduced 90-day mortality risk. CONCLUSION: Targeting specific PMs, especially diuretic avoidance and timely postdischarge outpatient visit, may improve outcomes in cirrhosis-associated AKI.