Association of patient characteristics with changes in estimated glomerular filtration rate before and after arteriovenous fistula/graft creation: a retrospective study

Abstract Background Planned creation of an arteriovenous fistula (AVF) or arteriovenous graft (AVG) before hemodialysis initiation is desirable to avoid unplanned hemodialysis and its associated adverse outcomes. Previous studies have reported a gradual decline in estimated glomerular filtration rate (eGFR) after AVF/AVG placement. However, the association of patient characteristics with this decline remains unclear. In this study, we aimed to investigate changes in eGFR from the initial nephrology referral to AVF/AVG creation and from AVF/AVG creation to hemodialysis initiation to elucidate differences by patient characteristics. Methods This single-center retrospective study included patients who underwent planned AVF/AVG creation and subsequently initiated hemodialysis between 2016 and 2023. Changes in eGFR were evaluated from the first nephrology referral to AVF/AVG creation and from AVF/AVG creation to hemodialysis initiation. Patients were divided into quartiles (Q1–Q4) based on ΔeGFR/year. Factors associated with eGFR decline were analyzed using logistic regression, and patient survival was assessed using the log-rank test. Results A total of 153 patients were included. The median ΔeGFR/year was significantly lower after AVF/AVG creation than before the procedure (3.0 versus 7.5 mL/min/1.73 m 2 /year, p < 0.001). Classification into Q1 was significantly associated with male sex and the absence of prior chronic kidney disease education hospitalization. Overall survival did not differ significantly according to the degree of change in eGFR slope. Conclusions In this retrospective study, eGFR decline was less steep after AVF/AVG creation. Male sex and the absence of prior CKD education hospitalization were associated with greater changes in eGFR slope. These findings should be interpreted cautiously because of the observational design and potential residual confounding.