Thus, patients with SBP who show DNI values greater than 5.0% should be managed very carefully. Third-generation cephalosporins have been recommended as the first line of antibiotic treatment for SBP. However, extendedspectrum empirical antibiotics such as carbapenems and piperacillin/tazobactam may be considered in the high-DNI group, as recent guidelines have recommended them for use in patients with nosocomial SBP. Although septic shock occurred more frequently in the high-DNI group, as is consistent with other studies, SIRS did not differ between the two groups. This is presumably because SIRS does not reflect well the infectious condition in cirrhotic patients due to factors such as baseline neutropenia and beta blocker use. Under this hypothesis, one can raise the question whether there exists any influence of neutropenia on the DNI value and its prognostic role. In a similar study, Pyo et al. investigated the role of DNI in the discrimination between disease flare-up and infection in patients with systemic lupus erythematosus patients in whom leucopenia are observed in some patients and leukocytosis are also frequently observed in other patients because of glucocorticoid usage, indicating that DNI reflects the proportion of immature granulocytes regardless of WBC count and can better reflect infection than WBC count which can be affected by other conditions without infection. Likewise, leucopenia is common also in cirrhotic patients. Therefore, DNI may be a useful indicator especially in cirrhotic patients with leucopenia. To confirm this novel suggestion, further prospective study should be performed. Recent reports have suggested that the MELD score could predict mortality in patients with SBP. However, in this study, the MELD score was unable to predict 30-day mortality in either univariate or multivariate Cox proportional hazard MLN4924 Metabolic Enzyme/Protease inhibitor analyses. This may be for several reasons. First, 80% of the patients enrolled in this study were categorized as Child-Pugh class C, so there may be no significant difference in underlying liver function among patients with advanced cirrhosis. Second, because MELD scores are commonly used as a 3-month mortality indicator in patients awaiting liver transplantation, it may not be possible to determine accurate associations between MELD scores and infection-related, short-term mortality. ARF has been known to be a risk factor for acute-on-chronic liver failure in recent studies, but in our study, it had no effect on 30-day survival. We believe that this phenomenon is a type 2 error caused by the small sample size. Although there is no statistical significance in the incidence of ARF between the two groups, the high DNI group, which was the independent predictor of 30-day mortality in our study, still showed a trend toward a higher incidence of ARF compared with the low DNI group. Therefore, we believe that ARF may affect 30-day mortality of SBP in a larger sample size. The connections among SIRS, multi-organ failure, and mortality have yet to be determined. Some studies have suggested that when inflammatory stress is superimposed on baseline cirrhosis. This results in an increased concentration of asymmetric dimethylarginine, an endogenous nitric oxide synthase inhibitor.