Abstract: PO1325
Vascular Access Type and Risk of Mortality and Hospitalization Among Elderly Hemodialysis Patients: A Target Trial Emulation Approach
Session Information
- Vascular Access
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 704 Dialysis: Vascular Access
Authors
- Lyu, Beini, University of Wisconsin-Madison School of Medicine and Public Health, Madison, Wisconsin, United States
- Chan, Micah R., University of Wisconsin-Madison School of Medicine and Public Health, Madison, Wisconsin, United States
- Astor, Brad C., University of Wisconsin-Madison School of Medicine and Public Health, Madison, Wisconsin, United States
Background
Evidence is mixed regarding optimal choices of incident vascular access type for elderly patients on hemodialysis (HD). Prior studies have primarily compared functioning arteriovenous fistula (AVF) to arteriovenous graft (AVG) and been limited to survival outcome. We used a target trial emulation approach and intention to treat (ITT) analyses to compare AVF versus AVG placement among elderly patients on HD.
Methods
Patients eligible for the target trial were those ≥67 years old at HD initiation, with no AVF/AVG placed before HD initiation, referred for AVF/AVG placement, and had AVF/AVG within 1 year after HD initiation. Patients would be randomly assigned to AVF or AVG and be followed right after AVF/AVG placement for 5 years. Outcomes including mortality, all-cause hospitalization, and cause-specific hospitalization (infection, cardiovascular disease (CVD), and vascular access (VA) related) within 6 months, 1 year, 3 years, and 5 years would be assessed. ITT analysis based on patients’ first AVF or AVG placed would be applied. We used USRDS data from 2010-2016 to emulate the target trial and propensity score (PS) matching to balance the groups’ characteristics.
Results
A total of 37,890 (out of 47,912) patients who had AVF/AVG placed within 1 year after HD initiation were included after PS matching. Among them, 28,847 (76.1%) had AVF placed and 9,043 (23.9%) had AVG placed. AVF was associated with lower risk of mortality over follow-up. Within 6 months after AVF/AVG placement, incidence of all-cause and VA-related hospitalization was significantly lower in the AVF group (RR 0.85 (95% CI: 0.78-0.93) for all-cause hospitalization; RR 0.68 (0.62-.74) for VA hospitalization), but not infection- or CVD-related hospitalization. AVF was associated with significantly lower incidence of all-cause and VA-, infection-, and CVD-related hospitalizations in longer follow-up time (RR 0.84 (0.82-0.87) for all-cause hospitalization within 3 years; RR 0.63 (0.59-0.67) for VA-related hospitalization within 3 years).
Conclusion
Our primary analyses found elderly patients on HD may benefit from getting an AVF compared to an AVG. We will further test whether these results hold true in patients within strata of age group, comorbidities, probabilities of AVF maturation, and life expectancy.