.png)
By Brian Rifkin, MD (@brian_rifkin)
Vascular access surveillance remains a controversial topic. Vascular access dysfunction is a complex process that is somewhat unpredictable, with major clinical implications. Sometimes the dialysis access will suddenly thrombose due to hypotension, external compression during sleep or thrombophilia, but often there are signs that an access is about to fail. The fundamental tenet of vascular access surveillance is that periodic assessment of dialysis access blood flow (Qa) can provide early detection of dysfunction and allow for timely percutaneous angioplasty (PTA) to prevent dialysis access complications and thrombosis. Advocates for surveillance believe that following Qa measurements over time lead to fewer critical access problems and less access loss. Declining access blood flow often leads to poor waste clearance or extended absence from critical dialysis therapy for access revision or thrombectomy. Opponents for screening argue that the benefits of Qa monitoring remain unproven. There are no consistent values for Qa that have the specificity and sensitivity to consistently detect critical stenoses. Qa values may vary significantly between dialysis treatments and are influenced by needle placement (with potential recirculation), serum protein levels, hydration status and cardiac output. Physiologic variation of repeated Qa measurements may exceed 20% between readings. Additionally, monitoring and interventions add expense with increased patient exposure to contrast and radiation. Finally, performing “unnecessary” PTAs has been associated with a controlled injury model that possibly leads to further fistula intima proliferation and stenosis.
The 2006 KDOQI Vascular Access Work Group previously recommended vascular access surveillance with preemptive angioplasty of stenosis to improve vascular access outcomes. Preemptive correction of luminal stenosis was recommended if stenosis exceeded 50% when the access flow was < 600 ml/min in AVGs and < 400-500 ml/min in AVFs without other signs of dysfunction. Unfortunately, this guidance was based upon many underpowered studies that showed neither benefit nor harm from surveillance strategies. In fact, recent meta-analyses did not show preemptive stenosis correction of functional vascular access improved access longevity; although results for AVFs were promising, existing evidence was insufficient to guide clinical practice and health policy. The updated KDOQI vascular access guidelines in 2019 suggest that positive dialysis access surveillance screening should not be intervened upon in the absence of clinical indicators. Clinical indicators include arm swelling or pain, difficulty with cannulation, prolonged bleeding, increased recirculation, inability to maintain prescribed pump settings and drops in URR or Kt/V that are not otherwise explained. In addition, physical assessment performed by experienced clinicians are sensitive for detecting stenoses and include a change in bruit/thrill pitch and quality, enlarging aneurysms, and non-collapsing vascular access with arm raise maneuvers. The new guidelines will lead to far fewer access interventions. Creating a consensus on values of Qa that are highly sensitive and specific for critical stenoses may once again bring access surveillance to the forefront. The current study attempted to investigate the optimal Qa to predict future access patency. They also examined whether weight-based adjustments were better at determining Qa thresholds for critical stenoses. No previous studies on optimal Qa thresholds evaluated Qa versus body size. Significant increases in Qa had been observed previously in overweight patients. It was hypothesized that a smaller body size requires less access flow, and a lower Qa cutoff value for intervention. This goes against the current convention that one-size-fits-all when measuring access blood flow.
The ultrasound dilution method (UDM) is one way to measure dialysis access Qa, and was utilized in this study. It requires specialized ultrasound sensors that are clipped onto the hemodialysis blood lines and transmit minute levels of ultrasound through the tubing wall into the bloodstream. The two transducers then pass ultrasonic signals back and forth through the access. The monitor derives a measure of the changes in time it takes for the wave of ultrasound to travel from one transducer to the other (“transit time”) resulting from the flow of blood. The velocity of ultrasound in blood (1560-1590 m/sec) is determined primarily by its blood protein concentration. A bolus of saline is injected and dilutes the blood protein concentration, reducing the ultrasound velocity. The arterial and venous sensors each register a dilution curve. Access recirculation, cardiac output and access flow velocity can be derived from dilutional curve data.
Valliant A, McComb K. Vascular Access Monitoring and Surveillance: An Update. ACKD. 2015 Nov;22(6):446-52.
For this study, clinical indicators of access dysfunction and Qa values were monitored in hemodialysis patients with radiocephalic AVFs. Patients with radiocephalic AVFs for >3 months were eligible for inclusion and evaluation. Qa surveillance was then performed by UDM. The development of access dysfunction necessitating percutaneous transluminal angiography was also analyzed. Patients with Qa < 500 ml/min, and without any clinical indicator of access dysfunction were included for outcome analysis. This would be consistent with 2006 KDOQI guidelines. Body weight was defined as the post-dialytic body weight at study entry.
52 patients were identified that had measured Qa < 500 ml/min (Fig. 1). These 52 patients received two Qa (Qa1 & Qa2) measurements during the follow-up period. 25 patients developed clinically significant stenosis or thrombosis within 3 months after the Qa1 measurement, necessitating PTA. After PTA, their access flow rate increased from a mean flow of 341.2 ± 105.3 ml/min to 934.0 ± 352.5 ml/min within the next 3 months (P < 0.001). In addition, 27 patients with asymptomatic AVFs were also subjected to a subsequent Qa2 measurements three months after Qa1. Three months after the Qa2, 17 patients remained asymptomatic, and 10 patients had PTA. The 17 patients who did not require PTA had higher Qa/actual body weight, Qa/BMI, Qa/body surface area, Qa/ideal body weight than those who experienced AVF functional loss. This data directly supports the theory that ideal thresholds for Qa vary by body mass, and should be considered accordingly when referring a patient for angiogram evaluation.
The study concluded that for patients with mature AVFs, Qa values suggesting stenosis were closely associated with weight. This suggests that a single Qa threshold for angiography may be too simplistic and that the optimal Qa threshold might be different based upon body mass. Multivariable logistic regression analysis revealed that a low Qa per ideal body weight was an independent predictor of AVF functional loss. Receiver operating characteristic curve analysis revealed that the best cutoff value of Qa was 7.1 times the ideal body weight. It was shown that by using weight based Qa assessments, the patients were able to avoid unnecessary invasive procedures. Additional large, randomized control trials are needed to further verify these findings and ultimately help create better guidelines for Qa monitoring thresholds.
Visual Abstract by @whatsthegfr