Blood stream Infections in chronic hemodialysis patients - characteristics and outcomes

Abstract

Abstract Introduction Bloodstream Infections (BSI) are a major cause of death and hospitalization among hemodialysis (HD) patients. The rates of BSI among HD patients vary and are influenced by local patient and pathogen characteristics. Modifications in local infection prevention protocols in light of active surveillance of BSI has been shown to improve clinical outcomes. The aim of this study was to further explore factors associated with BSI in a contemporary cohort of HD patients at a public teaching hospital dialysis center in Israel. Methods This was a retrospective cohort study of HD patients with a BSI in the years 2014 to 2018. The primary outcome was the occurrence of BSI. Secondary outcomes were to describe the causative pathogens of BSI, and to assess for risk factors for BSI, and mortality. Results Included were 251 patients. The mean age was 68.5 ± 13.4 years, 66.9% were male. The mean time from initiation of dialysis was 34.76 ± 40.77 months, interquartile range (IQR) 1-47.5 months and the follow up period of the cohort was 25.17 ± 15.9 months. During the observation period, 44 patients (17.5%) developed 54 BSI events, while 10 of them (3.9% of the whole cohort) developed recurrent BSI events. Gram-negative microorganisms caused 46.3% of all BSI events. 31.4% of these BSI were caused by resistant bacteria. In a multivariate logistic regression analysis, patients receiving dialysis through a central line had a significantly increased risk for BSI adjusted Odds Ratio (aOR) 3.907, p = 0.005, whereas patients’ weight was mildly protective (aOR 0.971, p = 0.024). Conclusions We noted an increased prevalence of gram-negative pathogens in the etiology of BSI in HD patients. Based on our findings, additional empirical antibiotics addressing gram negative bacteria have been added to our empirical treatment protocol. Our findings highlight the need to follow local epidemiology for implementing appropriate preventative measures and for tailoring appropriate empiric antibiotic therapy.