RESUMO
BACKGROUND: Catheter-restricted antimicrobial lock (AML) use reduces catheter-associated bloodstream infection (CA-BSI) in clinical trial settings, but may not be as effective in clinical settings and may increase bacterial resistance. DESIGN: Quality improvement report analyzed using a cross-sectional time series (unbalanced panel) design. SETTING & PARTICIPANTS: The study cohort comprised all prevalent adults treated with hemodialysis through a tunneled catheter for any, but not necessarily all, of the time from January 1, 2003, to June 30, 2006, in Manukau City, New Zealand (135,346 catheter-days, 404 tunneled catheters, 320 patients). QUALITY IMPROVEMENT PLAN: Catheter-restricted AMLs (heparin plus gentamicin) for all tunneled catheters from July 1, 2004. MEASURES: Repeated observations of CA-BSI, hospitalization, tunneled catheter removal, and death from CA-BSI analyzed by using generalized estimating equations with a single level of clustering for each tunneled catheter and patterns of bacterial resistance analyzed by using simple descriptive statistics. RESULTS: AML use was associated with reductions in rates of CA-BSI and hospitalization for CA-BSI by 52% and 69% for patients with tunneled catheters locked continuously with AMLs since their insertion compared with those with tunneled catheters that were not, respectively. AML exposure also was associated with a trend to increased gentamicin resistance amongst coagulase-negative staphylococci isolates, a pattern similar to that observed for BSIs in our general hemodialysis population in which tunneled catheters were not the source of BSI, but different from that in the general non-end-stage renal disease population in the region. LIMITATIONS: This is an uncontrolled observational study and cannot prove causality. The follow-up period of 18 months is longer than for other studies, but still too short to definitely answer whether AML use drives bacterial resistance. CONCLUSIONS: A change to use of AMLs may improve clinical outcomes; however, additional study of associated bacterial resistance is needed before AML use becomes standard care.