RISK FACTORS FOR ARTERIOVENOUS FISTULA FAILURE – AN OBSERVATIONAL STUDY IN A TERTIARY CARE CENTRE IN INDIA

Arteriovenous fistula (AVF) access is a lifeline for patients on hemodialysis given the lesser rates of infection and mortality for patients with AVFs as compared to those with central venous catheters and arteriovenous grafts. However, AVF failure can significantly influence patients’ prognosis. Hence, this study aims to identify potential risk factors for AVF failure in such patients.

A cross-sectional study was conducted among 205 adult inpatient CKD patients with AVF who seek maintenance HD at a tertiary care centre in South India. Patients were assessed using a semi-structured proforma. Demographic details, relevant medical history, baseline doppler ultrasound findings of the upper limb’s blood vessels, baseline blood investigations conducted prior to AVF creation, and AVF related data including the type of AVF, date of AVF creation and the duration of AVF use were collected. For patients with AVF dysfunction, the doppler ultrasound findings of the failed AVF were noted, if available, and the outcomes of these failed AVFs were also noted. Data were analyzed using SPSS version 25. Continuous variables were tested by Mann-Whitney U test while categorical variables by chi square test or Fisher’s exact test. To identify factors predictive of AVF failure, univariate and multivariate logistic regression analyses were done. Odds ratios were calculated, with 95% confidence intervals (CI). P values of < 0.05 were considered as statistically significant.

Out of the 205 study participants, 141 had proximal AVFs (69%) while 64 had distal AVFs. 72 of the 205 participants had failed AVFs (35%), of which 33 had primary AVF failure while 39 had secondary failure. AVF failure was found to have a significant association with history of dyslipidemia (p=0.001), systolic BP (p=0.013), diastolic BP (p<0.001), phosphorous (p=0.030), LDL cholesterol (p=0.008), presence of internal jugular catheter (IJC) at the time of AVF creation (p=0.029), radial artery diameter (p=0.003), proximal cephalic vein diameter (p=0.025), brachial artery peak systolic velocity (p=0.003), dialysis vintage (p<0.001), time since AVF creation (p=0.002) and duration of AVF use (p<0.001). Multivariate logistic regression detected a statistically significant association for AVF failure with four variables: history of dyslipidemia (p=0.020, OR (95% CI) = 24.29 (1.90 – 89.23)), systolic blood pressure (p=0.041, OR (95% CI) = 0.93 (0.89 - 90.97)), diastolic blood pressure (p=0.008, OR (95% CI) = 0.68 (0.51 – 0.87)) and phosphorus (p=0.033, OR (95% CI) = 1.90 (1.43 – 2.14)). 

Our study indicates that patients with AVF were more likely to develop secondary failure as compared to primary failure. Patients with failed AVFs were more likely to have older HD vintage and older age of AVF creation but lesser duration of use as compared to patients with patent AVFs. They were also more likely to have a history of dyslipidemia, low systolic and diastolic BP, high levels of phosphorous and LDL cholesterol, IJC present at the time of AVF creation, lesser radial artery diameter, lesser proximal cephalic vein diameter and lesser brachial artery peak systolic velocity. Thus, a comprehensive baseline evaluation of patients prior to AVF creation might be useful to identify those who may be at higher risk of experiencing AVF failure.