Tackling the Achilles' heel of hemodialysis.

More than 375,000 patients undergo long-term hemodialysis treatment in the United States, but the outcomes have remained abysmal, with the rate of death during the first year of hemodialysis therapy exceeding 20%. Although the development half a century ago of techniques for sustainable vascular access rendered long-term extracorporeal treatment feasible, vascular access remains the Achilles’ heel of hemodialysis. The current options include arteriovenous fistulas, synthetic grafts, and central venous catheters, with a clear hierarchy among these options. Native arteriovenous fistulas — simple anastomoses of forearm arteries and veins — yield the best outcomes. Among otherwise similar patients, those with functioning dialysis fistulas live the longest and have the fewest infectious complications. Such fistulas, however, need to mature for several weeks or months until they can accommodate the blood flow necessary for dialysis, and many fistulas never mature sufficiently for adequate use. Synthetic vascular grafts can be used in patients whose native vessels may not support a fistula. These grafts can be used sooner than fistulas but carry higher risks of infection. Least desirable of all the options is the implantation of a permanent, usually cuffed, central venous hemodialysis catheter. Although central hemodialysis catheters can be inserted quickly and are available for use immediately, they are associated with particularly high rates of infection, hospitalization, and death. These catheters are prone to partial or total occlusion, which may lead to inadequate dialysis and missed dialysis sessions. As a result, maintaining central venous catheters is costly and burdensome to the patient. All in all, most clinicians agree that the use of central venous hemodialysis catheters should be avoided whenever possible. There have been several initiatives aimed at increasing the number of patients in whom fistulas are used for hemodialysis. As a result, the use of fistulas has increased in recent years, predominantly replacing the use of synthetic grafts; in 2007, fistulas were used 55% of the time, and synthetic grafts 27% of the time.1 Unfortunately, 18% of patients still use central venous catheters, either because they start dialysis without a functioning peripheral vascular access or because all suitable peripheral vessels have been exhausted over the course of years of hemodialysis treatment and frequent vascular-access failures. In 2008, a central venous catheter was used in 74% of the patients in the United States who were undergoing hemodialysis for the first time as outpatients.2 Only 16% of patients had a maturing arteriovenous fistula or graft in place, strongly suggesting that more than half the patients initiating hemodialysis had to rely on cuffed central venous catheters for several months until a peripheral venous access could be established and would be available for use. The rates of associated serious infections during the first months of hemodialysis treatment have been exceedingly high — more than 200 hospitalizations for (systemic) vascular-access–related infections per 1000 patient-years in the first 6 months.3 For patients who must rely on a central venous catheter for hemodialysis treatment, the findings by Hemmelgarn et al. in this issue of the Journal provide important new evidence.4 A simple regimen in which recombinant tissue plasminogen activator (rt-PA) was used for sealing the two lumina of the dialysis catheter once a week (with heparin used for the other two treatments each week) was superior to a regimen in which heparin seals were used after all the treatments; the rate of access failure was halved,