Impact of practice changes on catheter-related exit-site and bloodstream infection rates in a Canadian hemodialysis center: A retrospective study

Background: Hemodialysis vascular access predisposes patients to exit-site infections (ESIs) and bloodstream infections (BSIs), resulting in significant morbidity and mortality. The objective was to characterize hemodialysis catheter-related (CR) ESIs and BSIs while considering potential factors associated with infection. Methods: The study period was selected to coincide with new CR-infection prevention measures at the midpoint. These included masking during exit-site care, using chlorhexidine-alcohol versus povidone-iodine antiseptic, administering cefazolin prophylaxis with central venous catheter (CVC) insertions, and reducing temporary CVC use for chronic hemodialysis starts. Data were collected retrospectively, including patient characteristics, hemodialysis history, CVC details, and CR-infections. Quarterly infection rates were calculated per 1000 CVC days, and potential factors associated with infection were investigated. Modeling was used to characterize infection rates and covariates over time. Results: Over 39 months, data for 267 patients, 499 CVCs, and 114,825 CVC days were captured. During the study period, there were 113 ESIs and 64 BSIs, with >80% of infections caused by gram-positive bacteria. ESI and BSI rates were 0.98 and 0.56 per 1000 CVC days, respectively. There were significant reductions in infection rates over time. The ESI rate dropped when new CR-infection prevention measures were introduced (p < 0.01), from a mean of 1.28 to 0.73 per 1000 CVC days (p = 0.003). The rate of BSI trended downward to a low of 0.10 per 1000 CVC days in the last quarter of the study. The BSI rates associated with temporary and permanent CVCs were 1.25 and 0.53 per 1000 CVC days, respectively (p = 0.1). There was a strong correlation between the declining BSI rates and declining temporary CVC use over time (rho = 0.73, p = 0.005). Conclusions: CR-ESI rates dropped significantly when new hemodialysis CR-infection prevention measures were introduced. CR-BSI rates declined over the study period, as did the use of temporary CVCs.