Objective: Document the incidence, epidemiology and microbiology of BSI-HD in Québec, to allow for provincial benchmarking.
Methods: The 39 HD units in Québec were invited to participate (all hospital-based). Definitions of BSI-HD are those of NNIS/NHSN prior to the recent changes. The denominators are per 100 month-patients(mp).Numerators and denominators were stratified according to venous access types : temporary catheter (TC), permanent catheter (PT), synthetic fistula (SF) or native fistula (NF). Pooled mean rates were obtained with 95% CI and compared using square root transformation test. Proportion comparisons were done with chi-square test..
Results: 26 units (66%) are now participating in the surveillance program (up to 89% of total dialysis patients). The global rates were respectively 0.49, 0.50 and 0.56 BSI-HD per 100 mp in 2007-08, 2008-09 and 2009-10. In 2009-10, stratified rates were NF =0.17, SF= 0.50, PC= 0.73 and TC=7.64 BSI-HD per 100 mp. BSI rate was 4 times higher with catheters (TC and PC) than fistulas (NF and SF) (p<0,0001). Proportion of venous access types varied: NF =40.6%, SF=5.8%, PC=52.8% and TC=0.8%. The most frequent microorganisms were S. aureus (48%) and coagulase-negative Staphylococcus (24%). 18% of S.aureus were MRSA.
For the 20 units that participated since 2007, there was a significant BSI-HD increase between 2007-08 (0.48 / 100 mp) and 2009-10 (0.61 /100 mp) attributable to a significant increase in NF related BSI-HD going from 0.06 /100 mp to 0.20 /100 mp respectively. In 2009-10, 21 of 26 BSI-HD episodes related to NF, were associated with the newly applied « button hole » technique.
Conclusions: This BSI-HD surveillance is useful for benchmarking and will provide hints to decrease the incidence of BSI-HD in the province of Québec. NF remains the best choice for hemodialysis venous access. The proportion of NF should be increased and reasons for its lesser use in Québec should be explored. Caution is advised with the « button hole» technique.