Pneumonia in patients with cirrhosis: risk factors associated with mortality and predictive value of prognostic models

Pneumonia in patients with cirrhosis: risk factors associated with mortality and predictive value of prognostic models

<h1>Pneumonia in patients with cirrhosis: risk factors associated with mortality and predictive value of prognostic models</h1>

Pneumonia in patients with cirrhosis: risk factors associated with mortality and predictive value of prognostic models

Introduction Cirrhosis is one of the most common causes of mortality worldwide especially in developing countries, with 1-year mortality ranging from 1 to 57% depending on stage [ 1 ]. Patients with cirrhosis require frequent medical support, which result in heavy healthcare burden. Cirrhosis not only is a chronic and progressive liver damage, but also involves in a multifactorial immune dysfunction, including uncontrollable cytokines secreting, low phagocytosis of the innate immune and abnormal reaction of T and B cells in pathogen stimulation [ 2 ]. Infectious diseases are common in patients with advanced cirrhosis and exert one of most important reasons for mortality. As reported previous, infectious complications increased mortality 4-fold in cirrhotic patients, 30% patients died within 30-day and another 30% died within 1 year after infection [ 3 , 4 ]. Pneumonia is a common infectious disease in patients with cirrhosis [ 5 , 6 ]. Importantly, in a infectious disease survey of 4576 cirrhotic patients, pneumonia had a 2.95-fold increase in 30-day mortality, the highest among all infection complications [ 7 ] In patients with care unit acquired pneumonia, cirrhosis worsen clinical outcome and increased 28-day mortality as high as 11-fold [ 8 ]. Cirrhosis and pneumonia impact each other in pathophysiology. On one hand, cirrhotic host was related to impaired both early and later neutrophil-mediated pulmonary killing of the organisms, making infection uncontrollable [ 9 ]; on the other hand, excessive inflammatory factors triggered by pneumonia often lead to rapidly deteriorate liver function and directly damp the anti-bacterial immunity, and further cause multi-organ damage [ 5 , 6 ]. Although pneumonia exhibits higher mortality in the patient with cirrhosis, few studies focused on the risk factors. To recognize the risk factors of mortality and to optimize stratification is critical for improving survival rate. The purposes of current study were to (1) investigate the epidemiology and outcome of cirrhotic patients with pneumonia; (2) examine independent risk factors for all-cause mortality within 30- and 90-day; (3) compare the value of prognostic models for cirrhotic patients with pneumonia. Patients and methods Study design and patients Five thousand seven hundred twenty seven adult cirrhotic patients (≥18 years) from a retrospective cohort referred between January 1, 2013 and January 1, 2015 in the First Affiliated Hospital of Zhejiang University School of Medicine (Hangzhou, China) were screened. Overall 203 cirrhotic patients with pneumonia were included in this study. Cirrhosis was diagnosed by (1) liver biopsy, (2) radiological evidence of liver nodularity and splenomegaly in patients with chronic liver diseases, (3) clinical evidence of signs of portal hypertension or hepatic decompensation (including ascites and hepatic encephalopathy (HE)) [ 10 ]. Pneumonia was defined as a new infiltrate focus on chest radiological exam and one or more symptoms as follows: respiratory symptoms (ie cough, chest pain, dyspnea), sign of infection (fever > 38 °C, temperature  12,000/ mm 3 or 1 mg/kg prednisone for>1 month)); (4) bone marrow or solid-organs transplantation; (5) patients with hepatocellular carcinoma or with other types of carcinoma (6) HIV-infected patients. Data collection The demographic and clinical information were collected: age, sex, smoking, alcohol abuse, etiologies and complications of cirrhosis, co-morbidity; history of operation and pneumonia within 3 months, laboratory parameters and radiographic findings, antibiotic therapy, intensive care unit (ICU) admission, severity models and prognosis. For all patients, the data were collected at diagnosis of pneumonia at admission or up to 5 days after the onset of pneumonia after admission. Community acquired pneumonia (CAP) were those present at admission or developed within the first 48 h after hospitalization. Nosocomial acquired pneumonia (NAP) were those diagnosed after 48 h of admission. Complications of cirrhosis (including ascites, HE, hepatorenal syndrome) were defined in patients according to criteria from the European Association for the Study of the Liver and International Ascites Club [ 11 ]. Systemic inflammatory response (SIRS) was diagnosed as at least 2 of the following terms: heart rate > 90 bpm; respiratory rate > 20 bpm; temperature of > 38 °C or  12,000/mm 3 or < 4000/mm 3 [ 12 ]. Bacteremia was defined as positive blood cultures. ACLF was defined as previous description in EASL-CLIF Acute-on-Chronic Liver Failure in Cirrhosis (CANONIC) study [ 13 ]. Appropriate empirical antibiotic use was considered as adequate that at least one antibiotic against to the isolated pathogen, according to susceptibility testing, or patients were improvement of clinical signs and laboratory exams of infection after 2–3 days’ empirical antibiotic therapy. Otherwise, the empirical antibiotic therapy was considered inappropriate. Prognostic models used to predict 30-day and 90-day mortality included: PSI, MELD, MELD-Na, CTP score, qSOFA and CLIF-SOFA score. PSI score was calculated as previously described [ 14 ]. CTP, MELD and MELD-Na are conventionally used to predicting the outcome of end-stage liver disease. Formula for MELD is: 9.6 × log e (creatinine, mg/dL) + 3.8 × log e (total bilirubin, mg/dL) + 11.2 × log e (INR) + 6.4 × (etiology: 0 for cholestatic or alcoholic cirrhosis, 1 for others) [ 15 ]. CTP score composed of ascites, HE, albumin, serum bilirubin and INR [ 16 ]. qSOFA includes three aspects: assigning one point for respiratory rate ≥ 22 breaths /min, systolic blood pressure ≤ 100 mmHg, or altered mentation [ 17 ]. ACLF-SOFA score was proposed to assess organ failure in ALCF patients by addressing six functional failures (hepatic, renal, cerebral, coagulatory, circulatory and respiratory) [ 13 , 18 ]. Statistical analysis Data were showed as mean ± standard deviations and discrete data were showed as median with the interquartile ranges (IQR). Frequency and percentage were presented for categorical data. Student’s test or Wilcoxon test were used to compare continuous variables in each group. Nominal variables were compared using chi-square test/ Fischer’s exact test. Differences were considered significant at the level of two-sided 0.05. Cox’s proportional hazard regression was used to exam risk factors of time-dependent death. Significant candidate variables ( p < 0.05) among bivariate analysis and possible variables were entered into a multivariate Cox’s regression by a backward-forward approach. Survival of the patients and subgroups was analyzed using Kaplan–Meier curve and pairwise Log-rank test. Receiver operating curve (ROC) were used to compare the predictive value of different prognostic scoring systems. The Youden index was used to identify the best cut-off point. Statistical analyses were performed using the SPSS software version 20 (IBM Inc., Chicago, IL, USA). Result Characteristics of population During the study, 5727 patients were diagnosed cirrhosis, 372 patients with cirrhosis and pneumonia. 169 patients were excluded as listed: 11 patients with liver or bone marrow transplantation; 3 patients with HIV positive; 4 patients with drug-induced immunosuppression; 19 patients lost to follow-up; 60 patients without completed information; 59 patients with cancer. 203 patients were enrolled in this study, 67 patients (31.0%) were non-survival and 136 patients (69.0%) were survival at the end of 90-day follow-up (Fig. 1 ). The median length of stay among cirrhotic patients with pneumonia was 20 days (Fig. 2 a). Among nosocomial acquired patients, the median length of stay was 10 days before occurrence of pneumonia (Fig. 2 b). Baseline characteristics and prognosis of the overall study cohort were depicted in Table 1 . Survival patients did not differ significantly from non-survival in relation to the causes of cirrhosis (virus 52.1% vs. 55.6%, alcohol 17.3% vs. 14.3%, others 35.0% vs. 36.5%). As shown in Table 1 , there were no significant differences between the survival and non-survival groups in sex, past medical history, smoking, antibiotic pre-treatment, and co-morbidities. In non-survival group, patients showed higher frequency of decompensated complication, ascites (90.5% vs. 72.9% p  = 0.005), hepatorenal syndrome (30.2% vs. 5.7% p < 0.001) and high grade (III-IV) HE (15.9% vs. 0.7% p < 0.001). Of note, bacteremia was more frequently observed in non-survived group (14.3% vs. 3.6% P  = 0.013). Significant differences were also observed in ICU admission (33.3% vs. 7.9% p < 0.001), SIRS (79.4% vs. 50.7% p < 0.001), and appropriate empirical antibiotic use (15.9% vs. 72.1% p < 0.001). Fig. 1 Flow chart of patients screening Fig. 2 ( a ) the length of stay among all patients; ( b ) the length of pneumonia occurrence among in-hospital cirrhotic patients Table 1 Comparison of clinical characteristics between survivors and non-survivors patients for 90-day follow-up Variables

Read More…

Enter your Email Address
Shree

Leave a Reply