Abstract
BACKGROUND: Candidemia is a life-threatening condition; however, the predictive markers for candidemia and mortality are inadequate in cirrhotic patients. This study was conducted to propose candidate predictors for the occurrence of candidemia and 30-day mortality in hospitalized cirrhotic patients with bloodstream infection (BSI) and review the related literature. METHODS: Cirrhotic patients with BSI between January 2011 and March 2020 were screened from the databank of a medical center and eligible patients were enrolled. Patients were separated into candidemia and bacteremia groups according to the results of blood cultures. Baseline characteristics, clinical presentation, and biochemistry data were collected at this time, as were microbiological data, medical management, use of antimicrobial agents, and outcome of the patients. The parameters and 30-day mortality were compared between candidemia and bacteremia groups. A combination of the MeSH terms and text terms related to candidemia and cirrhosis was searched in the electronic databases. RESULTS: Four hundred and sixty cirrhotic patients with BSI were enrolled. Thirty-five patients with candidemia (7.6%) were identified. Nosocomial infection, intensive care unit (ICU) admission, antibiotics exposure ≥14 days, white cell count >10 K/mm3, and model for end-stage liver disease (MELD) score >24 were associated with candidemia. The 30-day mortality was 65.7% in the candidemia group and 37.9% in the bacteremia group (p = 0.001). Nosocomial infection, ICU admission, hepatoma, hepatic encephalopathy, international normalized ratio ≥1.2, platelet ≤150 K/mm3, estimated glomerular filtration rate <60 mL/min/1.73m2, and MELD score >24 were associated with 30-day mortality. Six studies were identified. The results were consistent with our findings regarding low incidence of candidemia, and relevant risk factors are listed. CONCLUSION: Candidemia had low incidence but high mortality in hospitalized cirrhotic patients with BSI. New predictors were proposed for the occurrence of candidemia and 30-day mortality in these patients.