#6848 DIABETES AND HEMODIALYSIS ARTERIOVENOUS ACCESS CONSTRUCTION – A RETROSPECTIVE STUDY IN A SINGLE CENTER

Abstract

Abstract Background and Aims The preferred vascular access for patients with end-stage renal disease is native arteriovenous fistula (AVF). However, in some patients, given their co-morbidities, vascular access (VA) construction is not feasible. Diabetes Mellitus (DM) is described as one of the main factors associated with vascular access failure. The aim of this study is to characterize in a non-diabetic versus diabetic population the pre-surgical vascular mapping by doppler ultrasound (DUS), as well as analyze which VA could theoretically be constructed, what VA was effectively were constructed and its patency between the two groups. Method We present a retrospective study of a cohort of patients that were evaluated by DUS for pre-surgical planning and VA construction from 2018 to 2021. We collected demographic, clinical and DUS parameters. The feasibility for VA construction was classified in pre-surgical evaluation as possible, non-advisable or as borderline. The patency after surgery was observed until day 45. Results We analyzed a total of 355 patients. Almost half of the population had diabetes (n = 161, 45.4%). The demographics and clinical parameters are summarized in Table 1. Concerning pre-surgical vascular mapping, we observed that patients with diabetes had more severe calcifications than the non-diabetic group. Table 2 summarizes the doppler characteristics of right and left radial and brachial arteries in both groups. Regarding the feasibility of VA construction, the non-diabetic group was more likely to receive a radiocephalic fistula than the diabetic group (33% vs 23.6%, pvalue 0.12). Surprisingly, in our study the non-diabetic group had also more patients that could only sustain a prosthetic fistulas (15.5% vs 12.4%, pvalue 0.41). Diabetic patients showed more likelihood to construct proximal fistulaes (67.4% vs 56.8%, p = 0.06). In terms of what type of VA was effectively created, more radiocephalic fistulas were performed in the non-diabetic group (13.95% vs 7.5%, pvalue 0.17). In the diabetic group more brachiocephalic fistulas (65.6% vs 62.4%, pvalue 0.17) and more prosthestic fistulaes (19.4% vs 14.4%, pvalue 0.17) were constructed. Concerning VA patency at 45 days the diabetic group demonstrated higher rates but with no statistical significance (88.4% vs 83.2%, pvalue < 0.18). Only arterial hypertension, showed to be predictive for VA patency, in the univariate and multivariate analysis that included age, sex, diabetes and ethnicity (OR 0.3, pvalue <0.09 and OR 0.32, pvalue <0.019 respectively). Conclusion We found that diabetic patients had significantly more comorbidities (older age, obesity and peripheral artery disease). Regarding the construction of a VA, in diabetic patients there was a lower probability of creating a radiocephalic fistula but a similar possibility of constructing a proximal fistula. Besides that, after VA construction, no differences were found between the two groups in terms of vascular patency. Although diabetes status reduced the probability for radiocephalic fistulas, once VA construction was successful no differences in VA survival at 45 days was verified.