Utility of intra-access pressure monitoring in detecting and correcting venous outlet stenoses prior to thrombosis.

Vascular access thrombosis is a major problem for hemodialysis patients. Over 7.75 years, we performed intra-access venous pressure monitoring at zero dialyzer blood flow (VP0), correlated VP0 with access anatomy angiographically, and examined the effect of two levels of stenosis, 50% and > 65% luminal diameter reduction (% D) as selection criteria for referral and elective angioplasty or surgical revision upon access outcomes. Summary receiver outcome curves for absolute intra-access pressure (VP0) and intra-access pressure normalized for systemic pressure (VP0/systolic BP) were constructed to evaluate sensitivity and specificity and compared to recirculation and duplex color-flow Doppler. Access outcomes included thrombosis, revision, replacement, and angioplasty rates that were normalized per 100 patient years (100 pt-yrs). During the 7.75 year long study period totaling 832 patient-access years of risk, the percentage of prosthetic bridge grafts increased from 65% to 80%. SROC showed better sensitivity for normalized (VP0/systolic BP) than absolute intra-access pressure (VP0) in grafts. Recirculation had poor predictive power in prosthetic bridge grafts compared to VP0. Predictive power of recirculation was superior to VP0 in native arteriovenous fistulae. The angioplasty rate correlated inversely with the degree of luminal reduction used as selection criterion for referral for angioplasty or surgical revision. A strong inverse relationship between thrombosis rate and the angioplasty rate (R2 = 0.99) but not between thrombosis rate and the number of angiograms performed (R2 = 0.39) was noted. A consistent, yet evolving, intensive graft maintenance protocol produced a 70% decrease in the thrombosis rate, a 79% decrease in the access replacement rate, and an increase in the average age of patent usable vascular accesses from 1.97 to 2.98 years that was associated with a 13-fold increase in the angioplasty rate. We conclude that vascular access monitoring with VP0/systolic BP provides excellent selection criteria for angiographic referral. Intervention for stenotic lesions > 50% D using angioplasty or surgical revision markedly reduces thrombosis and access replacement rates.