A Systematic Review of Vascular Access Outcomes Trials in Hemodialysis Patients

Maintenance dialysis patients with a reliably functioning vascular access experience improved outcomes and overall well-being. However, according to Andrea K. Viecelli, MD, and colleagues, it remains challenging to establish and maintain a complication-free vascular access, and to reduce the need for recurrent interventions.

Complications related to vascular access account for ~20% of hospital admissions among patients with end-stage renal disease annually; such complications are associated with increased morbidity, morality, and healthcare costs. Vascular access has been called both the “lifeline” and the “Achilles’ heel” of hemodialysis. Patients concerned about the possibility of vascular access surgery and resulting complications experience increased stress and anxiety, making improving vascular access outcomes a high priority for patients, caregivers, and clinicians.

There have been numerous randomized controlled trials designed to improve outcomes related to vascular access in hemodialysis. Dr. Viecelli et al. recently conducted a systematic review to identify the scope and consistency of the outcomes and outcome measures used in contemporary trials and to examine the use of previously published standardized outcome definitions. The study design also included a secondary longer-term aim to “underpin strategies to prioritize outcomes, improve outcome reporting for vascular access complications, increase the value of future trials to inform evidence-based practice, and ultimately, help improve patient outcomes.” Review results were reported in the American Journal of Kidney Diseases [2018;71(3):382-390].

Following a search using Embase, the Cochrane Kidney and Transplant Specialized Register, and MEDLINE databases, the researchers identified 168 relevant trials involving 29,038 participants. Of the 168 trials, 113 were published (67%) and 55 (33%) were registered trial protocols. The published trials were conducted across 39 countries from North America, Latin America, Europe, Middle East, Asia, and Oceania.

The median duration of the trials was 12 (interquartile range [IQR}, 6-24) months and median sample size was 96 (IQR, 61-196) participants. Sixty-nine trials (41%) examined vascular access outcomes in arteriovenous fistulas alone; 53 (32%) in arteriovenous grafts; and 29 (17%) included various combinations of access type.

In total, 1426 outcomes measures were identified. The outcomes measures could be grouped into 23 outcomes, 65% (n=15) of which were predominately clinical, 17% (n=4) were surrogate, and 17% (n=4) were patient-reported.

The three most frequent outcomes were function (136 trials; 81%), infection (63 trials, 38%), and maturation (31 trials, 18%). The most common patient-reported outcome was pain, which was assessed in 19 (11%) trials. Five trials (3%) reported quality of life, two (1%) reported satisfaction with the vascular access, and one (<1%) reported needle phobia. The most commonly reported surrogate outcome was vessel characteristics, measured in 19 trials (11%).

Outcome Measures

Function was measured in 489 different ways and at 46 different time points. Because of the variety of ways function can be assessed, function was the most heterogeneous outcome. The most frequently used outcome measures to assess function included mean access blood-flow in milliliters per minute (37 trials, 27%) and number of thromboses (30 trials, 22%). Approximately one-third of outcome measures described patency, thrombosis, or stenosis.

Infection was reported in 136 different ways and at 20 different time points; 30 (22%) were consistent with criteria used by the US Centers for Disease Control and Prevention and/or other published definitions. The most commonly used outcome measure was the number of access-related infections (10% of trials). Catheter-related infections were assessed using 90 different outcomes measures; 64 referred specifically to systemic catheter-related infections; 11 referred to exit-site infections; and four to tunnel infection.

Measurement of maturation was conducted in 43 different ways and at 15 different time points; one-quarter of the trials referred to previously published criteria. The most common definition of maturation was the number of fistulas with a vein diameter of 6 mm and blood flow >600 mL/min and/or vein diameter of 4 mm and blood flow of at least 500 mL/min or as time to first cannulation (n=2, 6% trials for each outcome measure).

Nineteen trials examined patient-reported pain using 29 different outcome measures. Procedural pain was measured in two trials (10%); the majority of studies (n=15, 79%) measured pain during cannulation using one of 11 different methods to quantify pain.

Twenty-six trials reported access-related hematomas or bleeding from the access site; measurements were done in 29 different ways at 24 different time points. The most frequently used outcome measure was the number of hematomas (7 trials), followed by number of access-related bleeding episodes (4 trials). Vessel characteristics, the surrogate outcome, included measurements of vessel diameter, depth, and wall thickness; the outcome was measured in 26 different ways.

Limitations cited by the authors included the restricted sampling frame for feasibility and the focus on contemporary trials.

The researchers concluded by saying, “There is substantial variability and inconsistency in vascular assess outcomes and outcome measures reported in hemodialysis trials, with very little focus on patient-reported outcomes, making it difficult for clinicians, patients, and policy makers to make informed decisions. Clinicians are encouraged to judge current trial outcomes and measures based on the relevance to their patients and clinical practice.”

Takeaway Points

  1. Researchers in Australia conducted a systematic review to examine the scope and consistency of vascular access outcomes reported in contemporary hemodialysis studies.
  2. The review identified 168 relevant trials representing 29,038 participants; 1426 access-related outcomes were extracted and grouped into 23 different outcomes.
  3. The most common outcomes measured were function, infection, and maturation.