The Impact of Perioperative Ankle-Brachial Index and Clinical Status on Outcomes following Lower Extremity Bypass

Abstract

Ankle-brachial index (ABI) is a common method of graft surveillance after infrainguinal lower extremity bypass (LEB) surgery and is recommended by the Society for Vascular Surgery. Several studies failed to show the benefit of ABI surveillance but were limited by sample size, and the practice remains variable. We identified all patients who underwent LEB for occlusive disease from the Vascular Study Group of New England Registry (VSGNE) between 2003 and 2016. Postoperative changes were defined as follows: improvement for ABI >0.15 at discharge or clinical status improved (i.e., symptoms improved from rest pain to asymptomatic, and so forth), no change if ABI was within 0.15 or no change in clinical status, or worsened if ABI decreased >0.15 or clinical status deteriorated. We determined the independent effect of these changes on rates of mortality, reintervention, patency loss, amputation, and Major Adverse Limb Events (MALE-above ankle amputation, revision, thrombectomy, or lysis). In addition, we compared the practice of perioperative ABI to follow-up without ABI using propensity scores with inverse probability weights. We identified 7,994 patients undergoing their first intervention in the VSGNE, 2,251 of whom (29%) had both preoperative and discharge ABIs. Overall, 5,369 (67%) of the bypasses used vein, and 4,539 (57%) were femoropopliteal, with no difference in the rate of vein use or bypass type between those who had discharge ABIs and those who did not (P>0.05). Most bypasses were performed for chronic limb-threatening ischemia (59% in the ABI group, 65% in those without ABI data, P<0.01 for difference). At discharge, ABI remained stable in 22%, improved in 69%, and worsened in 9%, whereas clinical status remained stable in 12%, improved in 77%, and worsened in 12%. In univariate analysis, clinical status was associated with mortality, amputation, and MALE, but ABI change was only associated with mortality (all P<0.01). After multivariable adjustment, ABI change was no longer associated with mortality and remained unassociated with amputation and MALE (P>0.05), and the addition of ABI change to the models did not improve the fit of the model (likelihood ratio P>0.05). Forgoing perioperative ABI was associated with higher rates of patency loss (P=0.02), but not reinterventions (P=0.57), or untreated patency loss (P=0.17). A change in clinical status, but not a change in ABI, was associated with adverse outcomes after LEB. In this group of VSGNE patients with follow-up, perioperative ABI did not add incremental value to clinical status alone as a method of graft surveillance.