ORLANDO – Nosocomial infections in cirrhotic patients are associated with acute kidney injury and prolonged hospital stay, and thus adversely affect outcomes, findings from a Swedish population-based study suggest.
Of 344 patients diagnosed with cirrhosis between 2000 and 2010, 122 experienced a total of 230 bacterial infections; 29% were community acquired, 51% were health care acquired (HCA), and 20% were nosocomial. Most (70%) occurred in decompensated patients, and most patients (64%) used proton pump inhibitors. In-hospital mortality was 18%, Dr. Konstantina Sargenti reported at the annual Digestive Disease Week.
On logistic regression analysis, nosocomial/HCA infections, compared with community-acquired infections, were independently associated with PPI use (odds ratio, 2.07) and decompensated patient status (OR, 2.12). After researchers adjusted for confounders, nosocomial/HCA infections were not found to be associated with inpatient mortality (OR, 1.78) or systemic inflammatory response syndrome (OR, 1.18), but nosocomial infections alone were independently associated with hospital length of stay (OR, 23.34 per day) and acute kidney injury (OR, 2.82), which was defined by an increase of greater than 50% in serum creatinine, said Dr. Sargenti of Skane University Hospital, University of Lund (Sweden).
Study subjects – residents of an area in Sweden with a population of about 250,000 – all were diagnosed with cirrhosis during the study period. The patients were retrospectively evaluated, with all relevant hospitalization- and infection-related data extracted from medical records. They were then followed until death, transplantation, or the end of 2011, for a median follow-up of 50 months.
The groups of patients with community-acquired, HCA, or nosocomial infections did not differ significantly with respect to patient demographics, etiology of liver cirrhosis, or the presence of comorbidities or hepatocellular carcinoma. Length of stay was longer for those with HCA and nosocomial infections, but the three groups did not differ with respect to need for intensive care unit stay, in-hospital mortality, acute kidney injury, or systemic inflammatory response syndrome occurrence.
Also, PPI use was more frequent in the HCA and nosocomial infection groups, but the groups did not differ in their use of immunosuppressive or steroid use.
The results did not change in an analysis that included only the first bacterial infection for each patient, Dr. Sargenti said.
The most common types of infections were urinary tract infections in 22% of patients, spontaneous bacterial peritonitis in 19%, pneumonia in 14%, spontaneous bacteremia in 14%, and skin infections in 10%.
Nosocomial and HCA infections occur commonly in cirrhosis, but population-based data characterizing their occurrence and potential role in mortality or length of stay have been lacking, Dr. Sargenti said, noting also that while PPIs are known to increase the risk for infections, and acute kidney injury and systemic inflammatory response syndrome during an infectious episode are known to be associated with poor prognosis, it was previously unclear whether these conditions are more common in nosocomial and HCA infections.
The current findings suggest that most infections in a cirrhotic cohort are HCA or nosocomial infections, and that PPI use is an independent predictor of such infections. Nosocomial infections are a particular concern, as they appear to increase the risk of factors associated with poor outcomes, she concluded.
Dr. Sargenti reported having no disclosures.