PREDICTIVE FACTORS FOR PROSTHESIC ARTERIOVENOUS FISTULA OPTION BY DUPLEX DOPPLER ULTRASOUND VASCULAR MAPPING

Abstract Background and Aims Vascular access (VA) remains the lifeline for hemodialysis (HD) patients. Arteriovenous fistulas (AVF) are recommended over prosthesic arteriovenous fistula (PAF). However, the choice of the type of VA still reflects local practice differences and patient-specific demographic and clinical factors. Duplex Doppler ultrasound (DDU) has been shown to be useful in evaluation of both structural and functional aspects of the peripheral vessels, and is emerging as the preferred method for VA planning. Our aim was to find predictive factors for PAF creation in our population, i.e. when AVF was not feasible. Method We retrospectively analysed a cohort of chronic kidney disease patients who underwent DDU vascular mapping for preoperative planning of HD access at a tertiary referral centre from 2019 to 2020. All the exams were performed by the same DDU operator. Demographic, clinical and DDU characteristics were studied. Continuous variables were recorded as means (±SD) for normally distributed data or as medians (interquartile ranges) for non-normally distributed data. Comparisons were made using t tests or Wilcoxon rank sum tests as appropriate. Categorical variables were evaluated by frequency distribution and recorded as proportions. Comparisons were made using the x2 test. Unadjusted and adjusted multivariate logistic regression models were fitted to identify risk factors to PAF creation due to not being a good candidate for AVF. Results A total of 252 patients were included. The mean age was 65±16 years, 144 (57%) were male, 211 (84%) where white, 40 (16%) were black and 1 (0.4%) was asian. Two hundred and thirty seven (94%) patients were right-handed and 186 (74%) were being evaluated for the first vascular access. The majority had arterial hypertension (HT) (205 (81%)); 98 (39%) had Diabetes Mellitus (DM), 89 (35%) were current or past smoker and 72 (29%) were obese. Most patients were considered appropriate candidates for AVF (207, 82%), whereas 45 (18%) were assigned to PAF. Figure 1 compares the DDU’s arterial indexes of each group. In a multivariate logistic regression model adjusted for age, sex, HT, DM and obesity, black race was found to be a predictor of being a candidate for PAV creation (OR 2.46; CI 95% 1.05-5.71; p-value 0.036). Conclusion Our study revealed that black race is an independent factor for PAV creation even after adjusting for classic risk factors as age, DM or HT. Long known insidious factors, such as poor socioeconomic status or poor access to medical care, have been pointed as justifiers to this disparity. However, facing the PAF-associated risks, we believe that further work must be done to clarify potential involved anatomical factors and potential reversible factors. Those patients have been described throughout the literature as presenting with higher risk of VA failure and severe peripheral arterial disease. In parallel we found that radial and ulnar arteries diameters and radial pulse wave velocity (PWV) in DDU were significantly lower in PAV-submitted patients, suggesting distal arterial compromise and stiffness. Our work has some limitations: neither vein parameters nor VA outcomes were analysed. However, it introduces a relationship between the black race and worse arterial indexes, and their impact on the choice of type of VA, as they seem not to be good candidates for AVF.