Systemic Inflammatory Response Syndrome Predicts Acute Kidney Injury and In-Hospital Mortality Among Patients With Alcoholic Hepatitis

Abstract

Introduction: Alcoholic hepatitis (AH) is associated with high in-hospital mortality. Onset of acute kidney injury (AKI) in AH is associated with in-hospital mortality (IHM). Systemic inflammatory response syndrome (SIRS) frequently occurs in AH independent of presence of infection. Data are limited on the association of SIRS with AKI and IHM. We performed this study to test our hypothesis that SIRS among AH patients is associated with AKI and IHM. Methods: Medical charts of patients admitted to our center (2006-2013) with discharge diagnosis of AH (ICD-9 code 571.1) were reviewed to identify patients meeting criteria for AH diagnosis and total bilirubin ≥5 mg/dL. Results: A total of 116 patients (mean age 46 years, 66% male, 80% white, 52% with underlying cirrhosis, and 9% concomitant HCV) met the criteria for AH diagnosis. Ninety-five patients had severe AH with discriminant function (DF) ≥32 and/or hepatic encephalopathy. AKI was observed in 64 (55%) patients with 44 on admission (20 outside hospital transfer [OHT]) and 20 during hospital stay. Patients with AKI compared to without AKI differed for male gender (77% vs. 52%, p=0.006), OHT (43% vs. 19%, p=0.007), hepatic encephalopathy (49% vs. 21%, p=0.007), white cell count (15% vs. 9%, p=0.0005), infection (61% vs. 40%, p=0.03), mean DF (33% vs. 19%, p<0.001), and mean MELD (33% vs. 19%, p<0.0001). AKI occurred in 83%, 53%, and 37% of patients with no SIRS (n=33), uninfected SIRS (n=63), and infected SIRS (n=6), respectively, p=0.06 (Figure 1). A total of 74 patients (48 with AKI) received medical treatment for AH. After controlling for age, gender, SIRS, infection, treatment, and DF, male gender, infected SIRS, and DF predicted AKI with HR 2.4 (95% CI 1.3-4.9), 3.2 (95% CI 1.1-9.1), and 1.01 (95% CI 1.005-1.02), respectively. Patients with AKI had higher in-hospital and overall mortality (6% vs. 48%, p≤0.0001; and 29% vs. 65%, p=0.0001) with more frequent need for ICU care (46% vs. 17%, p=0.001). AKI remained a strongest predictor of IHM with over 8-fold higher mortality in presence of AKI: 8.4 (95% CI 1.8-39). Other predictors were female gender and DF. Conclusion: AKI is a significant event in AH patients and negatively impacts the patient survival. Presence of SIRS and infection are strong predictors of development of AKI. Prospective well designed studies are suggested to assess interventions at the time of hospital admission in reducing the risk of AKI among patients with AH.