부산대학교 도서관

Spontaneous bacterial peritonitis (SBP) is a common and severe complication in patients with cirrhosis and ascites. Intravenous albumin administration associated to antibiotic treatment has shown to decrease the incidence of renal failure and mortality in SBP. However, it is unclear whether it should be administered to all patients. Despite high rates of resolution with current antibiotic treatment associated to albumin treatment and improvements in general management of cirrhotic patients, in-hospital mortality related to SBP is still high, particularly in those with worse liver and/or renal function. The aim of this study was to evaluate the efficacy of albumin in a non-selected series of patients with SBP, to determine the independent predictive factors of in-hospital mortality at SBP diagnosis in high-risk patients treated with antibiotics and albumin and to create and validate a predictive model of mortality in these patients. We analysed all SBP episodes diagnosed during a 10-year period. Low-risk episodes (urea <11 mmol/l and bilirubin <68 µmol/l) were not treated with albumin, whereas high-risk episodes (urea >11 mmol/l and/or bilirubin >68 µmol/l) were or were not given albumin at the discretion of the attending medical team. To evaluate the efficacy of albumin we included 216 SBP episodes in 167 patients. Sixty-four episodes (29.6%) were low risk and not treated with albumin and 152 (69%) were high risk: 73 (48%) were treated with albumin and 79 (52%) were not. In the low-risk episodes, renal failure developed before SBP resolution was less frequent (4.7% vs 25.6%, p=0.001), in-hospital mortality was lower (3.1% vs 38.2%, p<0.001) and the 3-month probability of survival was higher (93% vs 55%, p<0.001) than in high-risk episodes. Considering only the high-risk group, there was a lower in-hospital mortality (29% vs 47%, p=0.02) and higher probability of survival at 3 months (63% vs 48%, p=0.02) among cases receiving albumin than in those not treated with albumin. To develop and validate a predictive model of in-hospital mortality in high-risk patients, we included 118 high-risk SBP episodes treated with antibiotics and albumin. In-hospital mortality was 33/118 (28%). The independent predictive factors of in-hospital mortality at SBP diagnosis were serum urea, blood leukocyte count, Child-Pugh score and mean arterial pressure. A predictive model including these four variables showed a discrimination accuracy (AUC) of 0.850, 95% CI 0.777-0.922. A cut-off point of 0.245 showed a sensitivity of 0.85 and specificity of 0.75. The in-hospital mortality was 28/49 (57.1%) in patients with a model value >= 0.245, and 5/69 (7.2%) in patients with a model value < 0.245 (p<0.001). The validation series included 123 patients with an in-hospital mortality of 33/123 (26.8%). The in-hospital mortality was 24/57 (42.1%) in patients with a model value >= 0.245, and 9/66 (13.6%) in those with a model value < 0.245. To conclude, our findings support the use of albumin treatment in patients with SBP and high risk of mortality but suggest that such therapy is not necessary in patients with low risk of mortality. We developed and validated a predictive model of mortality that includes serum urea, blood leukocyte count, Child-Pugh score and mean arterial pressure in high-risk SBP patients. These findings may help to identify patients who would benefit from additional therapeutic strategies.