HOME
Return to Table of Contents
PREDICTORS OF MORTALITY DUE TO PYOGENIC LIVER ABSCESSES: AN ANALYSIS OF US HOSPITALIZATIONS FROM 1998 TO 2004
L Meddings, RP Myers, AA Shaheen, KB Laupland, J Hubbard, G Kaplan
Division of Gastroenterology, University of Calgary, Calgary, AB
BACKGROUND: Pyogenic liver abscesses (PLA) are associated with significant mortality; however, few population-based studies have published outcomes for PLA in North America. We assessed the in-hospital mortality rate and determined independent predictors of mortality in a nationally-representative sample of US hospitals.
METHODS: We analyzed the 1998-2004 Nationwide Inpatient Sample that represents a stratified 20% random sample of all non-federal US hospitals. We used the International Classification of Diseases (ICD-9-CM) diagnosis codes to identify all discharges for PLA (572.0). We described the in-hospital mortality. We used logistic regression analysis to determine independent predictors of in-hospital mortality for PLA after adjusting for age, gender, insurance status, Elixhauser comorbidities, diabetes, cirrhosis, previous transplantation, and detection of a microorganism. Point estimates are presented as adjusted odds ratios (aOR) with 95% confidence intervals (CI).
RESULTS: Between 1998 and 2004 12504 discharges were reported for PLA. The median age was 62 years [intraquartile range (IQR) 48 - 73] and 58.7% were male. The in-hospital mortality rate was 6.0%. The majority of patients (76.5%) did not have an identifiable microorganism; the most common recorded bacteria were streptococcus species (8.9%), E. coli (5.0%), Klebsiella species (4.2%), and staphylococcus species (3.1%). Compared to patients under the age of 48 in-hospital mortality increased with age: 48 - 61 [aOR 1.34 (95% CI 1.10-1.63)]; 62-72 [aOR 1.77 (95% CI 1.36-2.30)]; and > 72 [aOR 2.46 (95% CI 1.84-3.30)]. Patients with cirrhosis [aOR 2.70 (1.95-3.75), without an identifiable microorganism (aOR 2.96 (2.30-3.80), and insured by Medicaid [aOR 2.14 (1.43-3.19)] were at increased risk for mortality. Gender, diabetes, prior history of liver transplantation, and the number of Elixhauser comorbidities were not predictive of death.
CONCLUSIONS: PLA was associated with significant in-hospital mortality. Several risk groups were identified including age, socioeconomic status, and liver cirrhosis. Lower mortality in patients who were culture positive suggests that targeted antimicrobial therapy may improve survival.