Venous access pressures and the detection of intra-access stenosis.

Venous pressure measured by the dialyzer is an unreliable measure of intra-access venous pressure. During dialysis and zero extracorporeal blood flow, intra-access venous limb pressure (VPd) was measured directly 401 times in 133 subjects using a high flow "in-line" three-way stopcock adjacent to the venous return needle. Subjects with systolic VPd/systolic blood pressure (BP) > or = 0.4, inadequate blood flow, or edema in the access extremity were referred for angiography. Percent diameter lumen reduction by a stenosis (%D) > 50% was considered hemodynamically significant. The authors did 138 angiograms. It was found that VPd/BP increased with %D in both ePTFE bridge grafts and native fistulae. Measurements of venous limb VP/BP taken at the time of dialysis and at the time of angiography did not differ (n = 55). On 80 occasions, accesses had significant stenoses. The overall sensitivity of VPd/BP in ePTFE bridge grafts was 91% and specificity 91%. False negative results occurred in seven of 24 native and eight of 114 ePTFE graft studies; 14 of 15 patients had arm swelling caused by central stenosis. Recirculation > 15% was more sensitive (71%) in detecting stenosis in native accesses than was intra-access pressure. It was concluded that VPd/BP > 0.4 is a useful, sensitive, and specific criteria for detecting synthetic bridge graft accesses at risk for thrombosis.