Vascular access for intensive maintenance hemodialysis: a systematic review for a Canadian Society of Nephrology clinical practice guideline.

ABSTRACT

BACKGROUND: Practices in vascular access management with intensive hemodialysis may differ from those used in conventional hemodialysis. STUDY DESIGN: We conducted a systematic review to inform clinical practice guidelines for the provision of intensive hemodialysis. SETTING & POPULATION Adult patients receiving maintenance (>3 months) intensive hemodialysis (frequent [≥5 hemodialysis treatments per week] and/or long [>5.5 hours per hemodialysis treatment]). SELECTION CRITERIA FOR STUDIES We searched EMBASE and MEDLINE (1990-2011) for randomized and observational studies. We also searched conference proceedings (2007-2011). INTERVENTIONS: (1) Central venous catheter (CVC) versus arteriovenous (AV) access, (2) buttonhole versus rope-ladder cannulation, (3) topical antimicrobial cream versus none in buttonhole cannulation, and (4) closed connector devices among CVC users. OUTCOMES: Access-related infection, survival, hospitalization, patency, access survival, intervention rates, and quality of life. RESULTS: We included 23, 7, and 5 reports describing effectiveness by access type, buttonhole cannulation, and closed connector device, respectively. No study directly compared CVC with AV access. On average, bacteremia and local infection rates were higher with CVC compared with AV access. Access intervention rates were higher with more frequent hemodialysis, but access survival did not differ. Buttonhole cannulation was associated with bacteremia rates similar to those seen with CVCs in some series. Topical mupirocin seemed to attenuate this effect. No direct comparisons of closed connector devices versus standard luer-locking devices were found. Low rates of actual or averted (near misses) air embolism and bleeding were reported with closed connector devices. LIMITATIONS: Overall, evidence quality was very low. Limited direct comparisons addressing main review questions, small sample sizes, selective outcome reporting, publication bias, and residual confounding were major factors. CONCLUSIONS: This review highlights several differences in the management of vascular access in conventional and intensive hemodialysis populations. We identify a need for standardization of vascular access outcome reporting and a number of priorities for future research.