When it comes to dialysis, one method of accessing the blood to clean it gets championed above the rest. But quite a few specialists say there's not enough evidence to universally support the treatment's superiority or to run down the other options.

"When we talk to [dialysis] patients in the clinic, we cannot address their profound question: 'Which access is better for me?' " says Dr. Pietro Ravani, an epidemiologist at the University of Calgary in Canada. "We just don't know, yet we are selling patients on a certain one."

Ravani is talking about guidelines that encourage doctors to pursue connections for dialysis known as arteriovenous fistulas. Research says hemodialysis patients with fistulas have a reduced risk of death, blood clots and infections compared with other access methods.

The connections require surgeons to stitch together an artery and vein, usually in the arm, to create a sturdier vein with greater blood flow. Patients are then pricked at the site of the fistula during each visit to connect to the blood-cleaning hemodialysis machine.

About 450,000 people in the U.S. are on dialysis.

Studies, like this one that was published in May, have shown patients with the fistulas had a lower risk of death (about a third less) when they start dialysis with fistulas rather than catheter connections.

But Ravani says not so fast. "The literature that is available and used to promote fistulas is biased," he says, adding there is no way catheters, an alternative to fistulas, are as deadly as some others have concluded.

Catheters are small plastic tubes, usually placed in a vein along the neck, chest, leg or groin, that can also be conduits for infection. Catheters are the go-to method for access to the blood when the kidneys suddenly fail and patients crash into dialysis, requiring emergency hospitalization and treatment. Fistulas can't be used for one to three months after an operation. Catheters can be used immediately.

Studies comparing these two access types and their mortality rates have only been observational, Ravani argues. That means researchers have looked at what happens to patients after doctors decided on their own how to treat patients. A randomized controlled trial that assigns patients to one treatment or the other and then collects information on what happens to them is necessary to ultimately prove the superiority of one method over another, Ravani says.

Patients with catheters, he explains, are usually pretty sick. But because it takes fistulas several months to develop before use, they are typically given to healthier patients who aren't in immediate need of dialysis.

"The very strong association between catheters and mortality could be related to how sick the patients were, not to the access type," Ravani says. "When you need to start dialysis urgently, it's because you're very sick so you use catheters, not fistulas. This makes it hard to determine if the poorer outcomes observed in patients with catheters are because of catheter or because they are already very sick."

For this same reason — serious illness — Ravani argues that patients with catheters succumb to infection more often than healthy patients with fistulas. If a healthy patient used a catheter, they wouldn't be as likely to contract an infection.

Nephrologist Swapnil Hiremath, at Ottawa Hospital in Canada, agrees that more research is needed to fully assess the value of fistulas. "The portrayal that fistulas are the ultimate access [for dialysis] and that if everyone has one, mortality rates will go down, is an exaggeration," Hiremath says. "You cannot go around blaming catheters; it's the nature of things that these patients are sicker and have a higher risk of death."

Hiremath adds that despite initiatives to increase the number of dialysis patients with fistulas, the treatment method is extremely difficult to provide to patients in the first place.

Roughly half of fistulas fail to mature, particularly in older individuals, and don't end up being used to access the blood, he says. Doctors then resort to catheters or another method to connect patients to dialysis machines. What's more, some 30 percent of patients completely reject the proposal of a fistula, despite explanations of their benefits.

And patients with fistulas can develop complications, such as heart failure, blood clots and swelling.

"To say that everyone who has a catheter should have a fistula, that's not easy," Hiremath says. "Doctors need to have an open mind, but unfortunately many people have already decided that fistulas are the best option."

Johns Hopkins University surgeon Dr. Mahmoud Malas, lead author of the recent paper on the advantages of fistulas, says Ravani's and Hiremath's criticism doesn't make sense to him.

Malas and his colleagues were behind an observational study showing patients starting dialysis with fistulas had lower risks of death. Despite the fact that he and his colleagues only reviewed existing numbers in the U.S. Renal Data System, Malas says they were able to minimize bias by matching the characteristics of patients with fistulas and catheters.

"If we saw a male patient with a catheter that was 40 years old who had diabetes and hypertension, we would find his exact match in a patient using a fistula," Malas explains. "Even with this matching analysis, you still see a much higher mortality rate for those on catheters."

"And our finding is not new, hundreds of prior studies have shown this difference," he adds.

Either way, Malas doubts a randomized trial could ever be carried out to truly compare those on fistulas and catheters. "Nobody would approve that trial," he says. "People will think it's unethical."

Ravani and Hiremath think differently. They are currently pursuing a randomized trial in Canada to tease out the differences between the two methods once and for all.

"For 40 years we have ignored this question with a randomized trial," Ravani says. "And until we have this answer, we cannot say fistulas are better."

...