Risk factors for recurrent bacteremia in adult patients with nontyphoid salmonellosis

Ron Bin Hsu, Robert J. Chen, Shu Hsun Chu

Research output: Contribution to journalArticlepeer-review

15 Citations (Scopus)


Background: This study sought to find the risk factors for recurrent bacteremia in adult patients with non-typhoid salmonellosis. Method Retrospective chart review. Result: Between September 1984 and December 2003, 235 adult (age ≥ 18 years old) patients with bacteremia with nontyphoid salmonellosis were admitted to our hospital. Among them, 130 patients (55%) had immunodeficiency, 31 patients (13%) had systemic lupus erythematosus, 26 patients (11%) had hematologic malignancies, 50 patients (21%) had solid organ cancers, and 39 patients (17%) had endovascular infections. Thirty-seven patients had recurrent bacteremia during the study period. Both univariate and multivariate analysis showed that immunodeficiency was the only predictor of recurrent bacteremia (odds ratio, 2.79; P = 0.013). The overall hospital mortality rate was 26%: 8% for patients with recurrent bacteremia and 29% for patients without recurrence. The independent risk factors of hospital death were old age, not recurrent infection, and solid organ cancers. Conclusion: Old age, systemic lupus erythematosus, malignancies, and immunodeficiency were common in adult patients with nontyphoid Salmonella bacteremia. The incidence of recurrent bacteremia was 16%. Immunodeficiency predisposed patients to recurrent bacteremia. Recurrent bacteremia was associated with a lower hospital mortality rate, however.

Original languageEnglish
Pages (from-to)315-318
Number of pages4
JournalAmerican Journal of the Medical Sciences
Issue number6
Publication statusPublished - Jan 1 2004
Externally publishedYes


  • Bacteremia
  • Nontyphoid Salmonella
  • Recurrence

ASJC Scopus subject areas

  • Medicine(all)


Dive into the research topics of 'Risk factors for recurrent bacteremia in adult patients with nontyphoid salmonellosis'. Together they form a unique fingerprint.

Cite this