Authors: Tz-Heng Chen, Yan-Ting Shiu, Timmy Lee, Chih-Yu Yang, Der-Cherng Tarng
Categories: Review Article, Flow surveillance, Hemodialysis vascular access, Noninvasive monitoring, Stenosis detection, Wearable devices
Source: Journal of the Chinese Medical Association : JCMA
Authors: Tz-Heng Chen, Yan-Ting Shiu, Timmy Lee, Chih-Yu Yang, Der-Cherng Tarng
A well-functioning vascular access is essential for delivering adequate hemodialysis in patients with end-stage renal disease. However, vascular access dysfunction, particularly stenosis and thrombosis, remains a leading cause of morbidity, repeated interventions, and hospitalization in this patient population. Vascular access monitoring and surveillance are designed to detect hemodynamically significant stenosis early, thereby reducing the risk of thrombosis and maintaining access patency. Evidence from meta-analyses and randomized controlled trials suggests that access blood flow (Qa)-based surveillance may lower thrombosis rates in arteriovenous fistulas (AVFs), while the benefit appears less consistent for arteriovenous grafts (AVGs). Consequently, most guidelines recommended incorporating Qa surveillance into routine clinical monitoring for AVFs, but not as a standard practice for AVGs. However, previous studies have notable limitations, including heterogeneous surveillance protocols and variable definitions of access dysfunction. More rigorously designed randomized controlled trials are needed to clarify the role of Qa surveillance and inform optimal strategies. Looking ahead, emerging technologies such as artificial intelligence and wearable devices for continuous monitoring hold promise for enhancing diagnostic accuracy, enabling earlier detection of dysfunction, and reducing the need for intervention rates. Integrating these innovations with standardized surveillance protocols and individualized patient risk stratification has the potential to improve vascular access longevity, reduce the healthcare burden, and improve outcomes in the hemodialysis population, although further validation is required.
Vascular access is the lifeline for patients with end-stage kidney disease (ESKD) undergoing maintenance hemodialysis. Inadequate vascular access function can lead to inadequate dialysis, increased morbidity, and higher healthcare costs.^1^ Common types of vascular access include arteriovenous fistulas (AVFs), arteriovenous grafts (AVGs), and central venous catheters (CVCs). Among these, AVFs are generally preferred because of their superior long-term patency, lower thrombosis and infection rates, and associations with improved patient survival and quality of life.^2^ However, AVFs are not always feasible due to anatomical or clinical constraints, leading to reliance on AVGs and CVCs. Despite advances in surgical techniques and access management, vascular access dysfunction remains a major clinical challenge, primarily driven by stenosis and thrombosis.^3^ Vascular access surveillance refers to the periodic assessment of vascular access using device-based methods or specialized tests beyond routine clinical examination. A widely used approach is measurement of access blood flow (Qa) to identify early hemodynamic changes suggestive of stenosis, enabling timely interventions through angioplasty or surgery.^4^ However, the clinical utility and cost-effectiveness of routine surveillance programs remain debated. Several studies have reported reductions in thrombosis rates for AVFs with surveillance methods, but evidence for AVGs is less consistent.
In this review, we examine the pathophysiology of vascular access dysfunction, describe key surveillance and monitoring modalities, and evaluate recent evidence regarding their impact on clinical outcomes. Additionally, we discuss future perspectives that may refine and personalize the management of hemodialysis vascular access.
Vascular access dysfunction in hemodialysis patients is primarily driven by stenosis and thrombosis, which compromise access patency and necessitate frequent interventions. The development of stenosis and thrombosis results from multiple mechanisms, including neointimal hyperplasia, endothelial dysfunction, inflammation and oxidative stress, shear stress, prothrombotic state, and fibrotic venous remodeling.
A central mechanism in the pathogenesis of vascular access stenosis is the proliferation and migration of vascular smooth muscle cells (VSMCs) from the medial layer into the intima. The accumulation of extracellular matrix (ECM) proteins further contributes to stenosis.^5^ Surgical trauma during AVF or AVG injures the endothelium and triggers the release of growth factors and cytokines, such as platelet-derived growth factor and transforming growth factor-beta (TGF-β). These mediators stimulate both proliferation and migration of VSMCs, leading to a neointimal thickening.^6,7^ Recurrent injury from hemodynamic shear stress, repeated needle puncture, angioplasty, or exposure to the bioincompatibility of AVGs can exacerbate neointimal hyperplasia and accelerate the progression of stenosis^8,9^.
Endothelial dysfunction is a critical factor in patients with ESKD and may be present even before vascular access placement.^10^ Endothelial dysfunction can result from multiple factors such as genetic predisposition, advanced glycation end products, hyperglycemia, high blood cholesterol, hypertension, obesity, diabetes, and smoking,^11^ which collectively create an imbalance between vasodilatory and vasoconstrictive substances produced by the endothelium.^12^ These alterations increase platelet aggregation, reduce nitric oxide (NO) production, elevate secretion of cytokines, chemokines, and adhesion molecules, and amplify reactive oxygen species generation.^13^
In ESKD, the accumulation of uremic toxins further disrupts the endothelium’s protective functions. Among these toxins, asymmetric dimethylarginine, an endogenous inhibitor of nitric oxide synthase (NOS), is frequently elevated, reducing NO bioavailability, thereby diminishing the vasoprotective capacity of the endothelium.^14^
Two NOS isoforms are typically expressed in endothelial inducible NOS (iNOS) and constitutive NOS (cNOS). The constitutive isoform plays a critical role in vascular vasodilation. Reduced production of NO derived from cNOS has been documented in advanced kidney disease, and contributes to impaired endothelium-dependent vasodilation.^15^
Inflammation and oxidative stress are common in chronic kidney disease and progressively worsen over time, contributing to vascular access dysfunction.^16^
Elevated cytokines such as tumor necrosis factor-alpha (TNF-α), TGF-β, insulin-like growth factor 1 (IGF-1), interleukin-1β (IL-1β), and interleukin-6 (IL-6) promote vascular remodeling and dysfunction.^17–21^ Furthermore, inflammatory cells are also major sources of matrix metalloproteinases (MMPs), which play a critical role in ECM degradation and smooth muscle cell migration. Excessive MMP activity accelerates neointimal hyperplasia and vascular remodeling, contributing to the failure of hemodialysis vascular access.^22,23^ Oxidative stress amplifies these processes through activation of nuclear factor kappa B signaling.^24^ Hemodialysis patients exhibit increased circulating levels of intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and monocyte chemoattractant protein-1 (MCP-1), which rise further during dialysis. This suggests that the dialysis procedure itself may exacerbate oxidative stress and inflammation, although the precise mechanisms remain unclear.^25^ Repeated needle punctures may exacerbate endothelial injury, facilitating platelet adhesion and release of MCP-1, VCAM-1, ICAM-1, IL-1β, and TNF-α. These cytokines amplify the inflammatory cascade, enhance cellular adhesion, and ultimately contribute to plaque or thrombus formation.^22^
Shear stress plays a significant role in the vascular remodeling of AVF creation. The abrupt increase in blood flow exposes venous endothelial cells to supra-physiological shear forces, triggering endothelial activation and structural adaptation.^26^ Over time, as the vessel dilates and flow stabilizes, shear stress normalizes in distal segments. However, endothelial cells near the anastomotic site continue to experience persistent wall shear stress (WSS).^27^ High shear stress has been suggested to promote the upregulation of MCP-1 and MMP-2, promoting neointimal hyperplasia.^28,29^ In these regions, excessive stress may also cause endothelial cell loss and expose a prothrombotic surface.^30^ Computational fluid dynamics analysis has demonstrated that anastomotic angles greater than 46.5° are associated with increased flow disturbances, reinforcing the importance of optimal surgical techniques to minimize shear stress-related complications.^31^
AVGs are also affected by shear stress, contributing to stenosis and graft failure through neointimal hyperplasia. Experimental models with arteriovenous polytetrafluoroethylene (PTFE) bypass grafts implanted in the canine femoral artery demonstrated that reduced WSS at the graft-vein interface enhanced intimal thickening.^32^ A porcine model in which PTFE grafts were placed between the iliac artery and vein and evaluated using serial phase-contrast magnetic resonance imaging (MRI) showed that elevated WSS at the venous anastomosis was associated with progressive intima-media thickening and luminal narrowing, confirming the pivotal role of shear stress in driving venous stenosis and graft failure.^33^ Similarly, canine studies comparing PTFE grafts of different diameters found that both excessively low and excessively high WSS promoted neointimal and pseudointimal thickening, whereas grafts exposed to intermediate shear stress demonstrated superior patency. These findings suggest the existence of an optimal shear stress range necessary for graft longevity.^34^
Patients with ESRD often exhibit platelet dysfunction characterized by impaired adhesion, aggregation, and secretion responses.^35,36^ Uremia alters platelet membrane glycoproteins, disrupts intracellular signaling pathways, and increases nitric oxide and prostacyclin production, all of which impair platelet-endothelial interactions.^37,38^ Despite these abnormalities, thrombotic events remain common in ESRD, reflecting a paradoxical prothrombotic state. Studies have consistently reported a high incidence of thrombotic events in CKD and uremic patients.^39,40^ Contributing mechanisms include hemodialyzer-induced platelet aggregation; endothelial abnormalities such as elevated plasminogen activator inhibitor-1, increased von Willebrand factor release, oxidative stress, and increased homocysteine levels.^16,41–44^ and coagulation abnormalities, including elevated fibrinogen concentrations, increased activity of clotting factors VII, VIII, and IX-XII, increased tissue factor (thromboplastin), elevated fibrinopeptide A levels, and reduced protein C activity.^45–49^ Platelet-derived microparticles (PMPs), which are released during platelet activation and provide a phospholipid surface for coagulation complex assembly, also play a key role in promoting thrombosis in these patients. Hemodialysis patients have significantly higher PMP levels than healthy controls, with even greater counts in uremic patients who experienced thrombotic events compared to those who did not.^50^ Notably, PMP levels are higher in recipients of erythropoiesis-stimulating agent (ESA) than in non-recipients, which may help explain findings from randomized controlled trials (RCTs) linking higher hemoglobin targets to increased thrombosis, possibly due to ESA-associated toxicity.^51,52^
Venous remodeling and AVF creation are critical determinants of access maturation. Recent evidence suggests that preexisting or postoperative intimal hyperplasia (IH) may not be the primary driver of AVF outcomes.^53,54^ Instead, postoperative fibrotic remodeling of the venous wall appears pivotal. A prospective study of patients undergoing two-stage AVF creation demonstrated that medial fibrosis after surgery was strongly associated with nonmaturation, whereas IH contributed to access failure only when accompanied by high medial fibrosis. This finding indicates that excessive fibrotic remodeling of the venous wall is a central risk factor for AVF nonmaturation, with the extent of medial fibrosis determining the stenotic potential of IH.^55,56^ Therapeutic strategies targeting ECM turnover, collagen deposition, and profibrotic signaling pathways such as TGF-β may therefore represent a promising approach to improve AVF maturation outcomes.^57^
The evaluation of hemodialysis vascular access relies on clinical monitoring and surveillance techniques to detect early dysfunction. Timely identification enables preemptive intervention to correct stenosis, prevent vascular access thrombosis, and enhance overall access patency and function.
Physical examination remains the cornerstone of vascular access assessment and is typically performed at every dialysis session and during routine outpatient visits. According to the Kidney Disease Outcomes Quality Initiative (KDOQI) 2019 Clinical Practice Guideline for Vascular Access,^4^ evaluation generally
Inspection:
Palpation:
Auscultation:
These approaches enable prompt recognition of infection, inadequate maturation or flow, possible stenosis, and other complications.^58^ In addition to physical examination, intradialytic assessments may also reveal clinical indicators of a significant vascular access lesion,
According to the KDOQI 2019 guideline, patients who consistently demonstrate these clinical indicators in the presence of AV access stenosis may reasonably undergo preemptive angioplasty to reduce the risk of thrombosis and access loss.^4^
While clinical monitoring is cost-effective and simple to implement, it depends heavily on the clinician’s skill and experience, making subtle changes difficult to detect. Advanced surveillance techniques, therefore, play an important complementary role in detecting vascular access dysfunction.
Surveillance of hemodialysis vascular access typically involves specialized techniques to measure Qa and detect stenosis, utilizing methods such as Doppler ultrasound and the ultrasound dilution technique. These approaches provide objective, reproducible assessments of vascular access function, allowing detection of progressive stenosis before clinical symptoms appear. In contrast to clinical monitoring, which relies on physical examination, surveillance requires specialized equipment and trained personnel.^59^
The ultrasound dilution technique is widely used to assess access recirculation and detect stenosis by measuring Qa. The principle involves injecting a saline bolus into the venous return line while using ultrasonic sensors placed on the arterial and venous bloodlines to detect dilution changes in the blood. The Transonic hemodialysis monitor (Transonic Systems Inc., Ithaca, NY) is the most commonly used device and has been validated as both reliable and reproducible for Qa measurements.^60,61^ Studies in adult and pediatric hemodialysis patients have demonstrated that the ultrasound dilution technique is a practical and reliable method for evaluating vascular access flow. Lower Qa values significantly correlated with stenosis and an increased risk of thrombosis.^62,63^
Doppler ultrasound combines Doppler flow assessment with grayscale imaging to evaluate vascular access. It provides detailed information on access flow velocity, stenosis location, and vessel anatomy, offering superior visualization of anatomic abnormalities compared to the ultrasound dilution technique.^64^ Previous studies comparing color Doppler ultrasonography with digital subtraction angiography have demonstrated high sensitivity and specificity in detecting access stenosis.^65,66^
The choice of surveillance technique depends on practical factors such as equipment availability, staff expertise, and specific clinical objectives. Table 1 compares clinical monitoring, the ultrasound dilution technique, and Doppler ultrasound in terms of measurement process, clinical utility, advantages, and limitations.
In general, the ultrasound dilution technique can be performed during hemodialysis sessions and provides rapid, quantitative flow measurements with minimal operator dependency.^59,67^ In contrast, Doppler ultrasound offers additional anatomical detail and precise localization of stenosis but requires more time, specialized equipment, and skilled operators for accurate performance and interpretation.^64,68^
A decline in Qa is a key indicator of vascular access dysfunction, often reflecting progressive stenosis and increased thrombosis risk. Monitoring Qa trends allows early detection of access deterioration and timely intervention to prevent complications. Several studies have demonstrated that maintaining Qa above 500 mL/min is associated with statistically significant reductions in thrombosis and access loss.^69^ However, the optimal Qa threshold may vary depending on vascular access type. AVFs generally function at lower baseline flow rates compared to AVGs. Evidence suggests that a Qa threshold of approximately 450 to 500 mL/min predicts stenosis in AVFs,^70^ whereas AVGs may require a higher threshold of around 600 mL/min, reflecting their lack of endothelium.^71^ Despite the widespread adoption of surveillance methods, clinical trials have reported mixed results regarding the impact of routine Qa surveillance on patient outcomes, prompting continued discussion and debate in the field.
Several RCTs have evaluated the effectiveness of routine Qa surveillance for hemodialysis vascular access. In the Hemodialysis Access Surveillance Evaluation (HASE) trial, 436 patients with AVFs or AVGs were randomized to standard care alone or standard care plus monthly ultrasound dilution surveillance.^72^ The surveillance group experienced significantly fewer thromboses per patient (0.12 vs 0.23; p = 0.012). However, rates of endovascular interventions (0.99 vs 0.98; p = 0.95) and central vein catheter placements (0.04 vs 0.05; p = 0.65) were similar, suggesting that Qa surveillance may reduce thrombotic events without increasing overall intervention needs. Another multicenter RCT followed 196 prevalent hemodialysis patients with native AVFs randomized to either standard monitoring methods alone or standard monitoring plus Qa surveillance (combined Doppler ultrasound and ultrasound dilution technique) protocol conducted every 3 months.^73^ Patients with Qa <500 mL/min or >25% decrease underwent fistulography, surgery, or more frequent monitoring. After 1 year of follow-up, the Qa surveillance group had a significantly lower thrombosis rate (0.02 vs 0.10 thromboses per patient-year at risk; p = 0.03) and a higher assisted primary patency rate, although non-assisted primary and secondary patency rates did not differ significantly.
By contrast, an RCT of 126 patients with AVGs found that supplementing clinical monitoring with regular ultrasound surveillance increased preemptive angioplasties (1.05 vs 0.64 events per patient-year) but did not reduce thrombosis (0.67 vs 0.78 per patient-year; p = 0.37). The median graft survival was also similar between the two groups (38 vs 37 months; p = 0.93).^74^ These disparate findings suggest that AVFs may benefit more from Qa surveillance than AVGs. A summary of key RCTs is presented in Table 2.
Several meta-analyses have investigated whether routine Qa surveillance reduces vascular access thrombosis in hemodialysis patients, with a particular attention to differences between AVFs and AVGs. In most analyses, control groups were managed with standard clinical monitoring methods such as dynamic or static venous pressure measurement, physical examination, and intradialytic assessments. By contrast, the intervention arms used more direct flow-based assessments, such as ultrasound dilution or Doppler ultrasound, to measure Qa.
Muchayi et al^82^ pooled data from seven RCTs (727 participants) and found a modest but non-significant reduction in thrombosis risk (pooled risk ratio [RR] 0.87; 95% confidence interval [CI], 0.67-1.13). Subgroup analysis suggested benefits in AVFs (RR, 0.64; 95% CI, 0.41-1.01) but no effect in AVGs (RR, 1.06; 95% CI, 0.77-1.46). Hwang et al^83^ reviewed nine RCTs (990 hemodialysis patients), of whom 658 had AVFs and 332 had AVGs. Overall, Qa surveillance did not significantly reduce thrombosis (RR, 0.78; 95% CI, 0.55-1.11; p = 0.17). Subgroup analysis again showed a protective effect in AVFs (RR, 0.56; 95% CI, 0.35-0.92; p = 0.02), but not in AVGs (RR, 1.10; 95% CI, 0.67-1.82; p = 0.70). Ali et al^84^ analyzed 10 RCTs comprising 1430 patients and an overall RR of 0.73 (95% CI, 0.55-0.98), favoring Qa surveillance across all access types. This benefit was primarily observed in AVFs (RR, 0.55; 95% CI, 0.33-0.89) while AVGs showed no significant improvement (RR, 0.92; 95% CI, 0.65-1.29).
Taken together, these meta-analyses suggest a consistent benefit of Qa-based surveillance in reducing thrombosis for AVFs, whereas evidence for AVGs remains inconclusive. However, prior randomized trials and meta-analyses in this field share several limitations. First, there is high clinical heterogeneity; different studies used varied surveillance modalities (eg, Doppler ultrasound vs flow dilution techniques) and inconsistent intervention thresholds, making direct comparisons difficult. Second, referral criteria were not standardized across RCTs. Third, many studies omitted patient-centered outcomes such as patient quality of life and health economic measures. Finally, the HASE trial did not adjust for its cluster-randomized design, likely underestimating variance and inflating statistical significance. To address these gaps, the ongoing FLOW trial (Maastricht University) is designed to provide more definitive evidence on the effectiveness of routine Qa-based surveillance.^85^ This double-blind, multicenter, individually RCT compares a strategy of routine monthly flow-based surveillance, with preemptive intervention triggered by Qa <500 mL/min, against a strategy in which intervention occurs only when clinical indicators of dysfunction are present. The study incorporates robust statistical methods, includes patient-reported outcomes, safety endpoints, and cost-effectiveness analyses, and uses a hybrid parallel-crossover design to preserve blinding after interventions. Once completed, FLOW may provide robust evidence to guide the use of Qa surveillance in both AVFs and AVGs.
Several observational cohort studies have evaluated the impact of Qa surveillance on vascular access outcomes in hemodialysis patients. In a multicenter, prospective study of 243 vascular accesses (192 AVFs and 51 AVGs), Giannikouris et al^86^ investigated the effect of routine Doppler ultrasound in maintaining patency. Primary unassisted patency rates at 12, 24, and 36 months were 83%, 75%, and 72%, respectively, while secondary patency rates were 93%, 88%, and 83%. Although AVGs demonstrated lower patency than AVFs, they still achieved acceptable outcomes under the ultrasound protocol.
Corrective procedures guided by ultrasound findings, rather than clinical indicators alone, also contributed to sustained patency. In a retrospective study by Wu and Lin,^87^ 397 chronic hemodialysis patients (336 AVFs, 61 AVGs) combined Qa surveillance with thorough physical examination. The study assessed absolute thresholds (<400 or <500 mL/min) and relative thresholds (Qa <1000 mL/min and a ≥25% decline) for predicting stenosis. Predictive accuracy was 91.5% for AVFs and 72.1% for AVGs, with the relative threshold offering no additional benefit. These findings suggest that integrating Qa surveillance models with routine clinical monitoring improves early detection and treatment of stenosis in hemodialysis patients.
Another center implemented a proactive Doppler ultrasound evaluation protocol 4 to 8 weeks after AVF creation.^88^ Hemodynamically significant lesions were detected in 62 AVFs (40%), although only 11% showed clinical abnormalities. Early detection allowed prophylactic interventions for subclinical lesions, improving fistula maturation rates from 25% to 70% among those treated. Collectively, these cohort findings reinforce that ultrasound-based surveillance can complement clinical monitoring by identifying at-risk accesses earlier and supporting timely interventions.
Several nephrology societies have issued guidelines on hemodialysis vascular access surveillance, offering differing perspectives on the role of ultrasound dilution technique, Doppler ultrasound, and other techniques in addition to routine clinical monitoring.
The UK Renal Association advocates systematic observation combined with advanced surveillance to help predict and prevent access failure without distinguishing between AVFs and AVGs.^89^ By contrast, the KDOQI 2019 guideline concludes that there is insufficient evidence to recommend routine Qa surveillance, pressure monitoring, or imaging for stenosis beyond clinical monitoring to improve AVF patency, and it explicitly advises against routine Qa surveillance for AVG.^4^ Overall, most international guidelines emphasize clinical monitoring as the cornerstone of vascular access evaluation, with Qa surveillance reserved as a supplementary tool in selected circumstances. Table 3 summarizes the key recommendations from major nephrology societies.
Advancements in data analytics, device technology, and personalized medicine are refining approaches to vascular access surveillance, enhancing both precision and clinical applicability. Machine learning is increasingly used to support real-time monitoring for early detection of access dysfunction, while novel wearable sensors enable continuous, noninvasive assessment of vascular access status. In addition, remote monitoring platforms are expanding the feasibility of at-home surveillance, reducing reliance on in-clinic resources and enabling earlier intervention. Collectively, these innovations hold the potential to improve patient outcomes, but still require further validation in real-world clinical settings.
Recent progress in artificial intelligence (AI) and machine learning has spurred the development of predictive models that integrate clinical data (eg, demographic factors, comorbidities) with various surveillance metrics. By analyzing large multicenter datasets, these algorithms can detect early, often subtle indicators of stenosis. Several proof-of-concept studies illustrate that the potential of AI-driven techniques can enhance vascular access management, for instance, by recording acoustic signals from AVFs with digital stethoscopes and converting these recordings into spectrograms or Mel spectrograms for convolutional neural network analysis. These models can then distinguish normal from stenotic flow patterns, quantify stenosis severity, and even predict 6-month primary patency.^93,94^ Furthermore, in resource-limited settings, such automated screening methods can enhance patient safety by enabling early identification of high-risk vascular access cases and supporting self-screening strategies that reduce the burden on clinical workload.^95^
Meanwhile, other groups have used routine flow and pressure parameters, such as dialysis blood flow rates, dynamic venous or arterial pressures, and recirculation percentages, in machine learning models. By tracking these activated variables over time, algorithms can detect subtle changes in vascular access function before thrombosis occurs, potentially reducing the need for invasive interventions and improving long-term patency.^96,97^ In contrast to traditional approaches, which depend heavily on clinician expertise, machine learning leverages routinely collected data to predict dysfunction and may support more efficient, risk-based surveillance strategies.
Wearable monitoring technologies have emerged as a promising approach for noninvasive, real-time assessment of vascular access function in hemodialysis. Recent studies describe external sensors capable of continuously measuring AVF flow rates and detecting hemodynamically significant stenosis.^98^
In one study of 128 patients, a small external device was positioned over the AVF for 1 week, enabling remote data capture through a wireless transmission. The wearable system reliably identified flow rates below 1000 mL/min, with reported sensitivity approaching 100% and specificity at 75%, and accurately detected stenoses greater than 50%.^99^ Another group developed a “soft thermal sensor” that uses a miniature heater to induce minimal skin warming and then measures subtle surface temperature changes across a wide flow range (0-800 mL/min).^100^ The flexible devices, integrated with wireless data transmission, can be placed directly on the skin or surgically exposed vessels, providing continuous feedback. Collectively, these findings indicate that wearable sensors, coupled with remote monitoring platforms, may allow earlier detection of access dysfunction and potentially facilitate timely intervention for stenosis or thrombosis.
Vascular access dysfunction remains a major challenge in hemodialysis, with stenosis and thrombosis as the primary causes of access failure. Clinical monitoring continues to serve as the cornerstone of assessment, while device-based surveillance methods, such as the ultrasound dilution technique, Doppler ultrasound, and emerging analytic models, provide additional tools for earlier detection and intervention. Current evidence suggests that Qa surveillance confers greater benefits in AVFs than in AVGs, though the optimal strategy remains uncertain. Emerging technologies, including artificial intelligence, wearable sensors, and remote monitoring systems, show promise for continuous assessment and early detection. However, existing evidence is preliminary and requires validation in larger, real-world clinical studies. A personalized, risk-based approach that integrates established techniques with novel innovations may improve vascular access care. Further research will be essential to confirm their effectiveness, cost-efficiency, and long-term impact on these strategies.
The authors acknowledge the financial support for research purposes by the “Yin Yen-Liang Foundation Development and Construction Plan” of the School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan (113Q158505); the National Science and Technology Council, Taipei, Taiwan (112-2314-B-A49-059-MY3, 112-2811-B-A49A-039, 112-2321-B-A49-020, 114-2314-B-A49-069-MY3, and 114-2314-B-075-005); Taipei Veterans General Hospital, Taipei, Taiwan (V111C-155, VGHUST111-G6-7-2, V113B-021, and V114C-070); Szu-Yuan research foundation of internal medicine; and the “Center for Intelligent Drug Systems and Smart Bio-devices (IDS^2^B)” from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan. The funders had no role in the study design, data collection, analysis, interpretation, or manuscript writing.