Development of a decision tree to streamline infrainguinal vein graft surveillance

O. M. B. McBride (Lead / Corresponding author), R. Mofidi, G. D. Griffiths, A. R. Dawson, R. T. A. Chalmers, P. A. Stonebridge

    Research output: Contribution to journalArticlepeer-review

    10 Citations (Scopus)


    INTRODUCTION: Duplex ultrasound (DU) remains the gold standard for identification and grading of infrainguinal vein graft stenosis. However DU based graft surveillance remains controversial. The aim of this study was to develop a decision tree in order to identify high-risk grafts which would benefit from DU based surveillance.

    MATERIALS AND METHODS: Consecutive patients undergoing infrainguinal vein graft bypass were enrolled in a DU surveillance program. An early postoperative DU was performed at a median of 6 weeks (range: 4-9). Based on the findings of this scan and four established risk factors for graft failure (diabetes, smoking, infragenicular distal anastomosis, revision bypass surgery), a classification and regression tree (CART) was created in order to stratify grafts in to grafts which are at high- and low-risk of developing severe stenosis or occlusion. The accuracy of the CART model was evaluated using area under receiver operator characteristic curve (ROC).

    RESULTS: Of 796 vein graft bypasses performed (760 patients), 64 grafts occluded by the first surveillance visit, 732 vein grafts were entered into surveillance programme. The CART model stratified 299 grafts (40.8%) as low-risk and 433 (59.2%) as high-risk grafts. 126 (17.2%) developed critical vein graft stenosis. Overall 30-month primary patency, primary assisted and secondary patency rates were 76.2%, 83.6% and 85.3%, respectively. The area under ROC curve for the CART model was 0.88 (95% CI: 0.81-0.94). Primary graft patency rates were higher in low-risk versus high-risk grafts (Log rank=186, P<0.0001). Amputation rates were significantly higher in the high-risk grafts compared with low risk ones (Log rank= 118, P<0.0001).

    CONCLUSIONS: A clinical decision rule based on readily available clinical data and the findings of significant flow abnormalities on an early postoperative DU scan successfully identifies grafts at high-risk of failure and will contribute to safely improving the efficacy of infrainguinal vein graft surveillance services.

    Original languageEnglish
    Pages (from-to)182-189
    Number of pages8
    JournalAnnals of Vascular Surgery
    Early online date25 Jun 2016
    Publication statusPublished - Oct 2016
    Event47th Annual Scientific Meeting of the Vascular Society of Great Britain and Ireland - Manchester, United Kingdom
    Duration: 28 Nov 201230 Nov 2012


    Dive into the research topics of 'Development of a decision tree to streamline infrainguinal vein graft surveillance'. Together they form a unique fingerprint.

    Cite this