PREDICTING HEMODIALYSIS ACCESS FAILURE WITH COLOR-FLOW DOPPLER ULTRASOUND

Color flow doppler ultrasound examination of the hemodialysis access w as conducted in 2,792 hemodialysis patients to evaluate its value in p redicting hemodialysis access failure. After baseline assessment of va scular access function with clinical and laboratory tests including co lor flow doppler evaluation these patients were followed for a minimal of 6 months or until graft failure occurred (defined as surgery or an gioplasty intervention, or graft loss). The patient demographics and v ascular accesses were typical of a standard hemodialysis patient popul ation. On the day of the color flow doppler examination systolic and d iastolic blood pressure, hematocrit, urea reduction ratio, dialysis bl ood flow, venous line pressure at a dialysis blood flow of 250 ml/min, and access recirculation rate were measured. At the conclusion of the study 23.5% of the patients had access failure. Case mix predictors f or access failure were determined using the Cox Model. Case mix predic tors of access failure were race, non-white was higher than white (p < 0.005), younger accesses had a higher risk than older accesses (p < 0 .025), accesses with prior thrombosis had a higher risk of failure (p = 0.042), polytetrafluoroethylene (PTFE) grafts had a higher risk than native vein fistulae (p < 0.05), loop PTFE grafts had a higher risk t han straight PTFE grafts (p < 0.025), and upper arm accesses had a hig her risk than forearm accesses (p = 0.033). Most significant, however, was decreased access blood flow as measured by color flow doppler (p < 0.0001). The relative risk of graft failure increased 40% when the b lood flow in the graft decreased to less than 500 ml/min and the relat ive risk doubled when the blood flow was less than 300 ml/min. This st udy has shown that color flow doppler evaluation, quantifying blood fl ow in a prosthetic graft, can identify those grafts at risk for failur e. In contrast, color doppler volume flow in native AV fistulae could not predict fistula survival. This technique is noninvasive, painless, portable, and reproducible. We believe that preemptory repair of an a natomical abnormality in vascular access grafts with decreased blood f low may decrease patient inconvenience, associated morbidity, and asso ciated costs.