Relation between coronary artery disease, baseline clinical variables, revascularization mode, and mortality☆

https://doi.org/10.1016/S0002-9149(00)01126-7Get rights and content

Abstract

The Coronary Angioplasty vs. Bypass Revascularisation Investigation (CABRI) trial comparing percutaneous transluminal coronary angioplasty (PTCA) with coronary artery bypass grafting did not show a difference in mortality with either procedure. Nonrandomized studies suggest that coronary artery disease (CAD) severity and distribution influences outcome. In the present study we explored the effect of prerevascularization CAD on 1-year mortality in the CABRI population, while adjusting for other baseline variables. Of the 1,054 patients recruited, there were sufficient angiographic results to derive the CAD scores in 974 (92.4%). Of these 974, there were 32 deaths. A number of CAD scores, both weighted for proximal disease (Duke and Leaman) and nonweighted, were used. These scores were then cross-tabulated against mortality. Demographic and clinical variables were also cross-tabulated against mortality and used to derive an initial logistic regression model to predict mortality. The effect of adding each of the CAD scores to this initial model was then assessed. After inclusion of the CAD scores, the best model was: (1) presence of peripheral vascular disease (odds ratio [OR] 3.89, p = 0.0025), (2) previous cerebrovascular accident (OR 2.86, p = 0.043), (3) older age (OR 1.05, p = 0.039), (4) a higher Duke score (OR 2.84, p = 0.0061), and (5) having undergone PTCA (OR 2.12, p = 0.047). In the CABRI population, adjustment for baseline variables, including prerevascularization CAD, revealed significantly higher mortality in those who underwent PTCA than in those who underwent coronary artery bypass grafting.

Section snippets

Study population

The present study population was that of the CABRI trial. Over 53 months starting in July 1988, 1,054 patients in 26 cardiothoracic centers throughout Europe were recruited: 542 underwent PTCA, of whom 522 had initially been randomized to PTCA, and 22 to CABG; 493 underwent CABG, of whom 478 had initially been randomized to CABG and 15 to PTCA. The inclusion and exclusion criteria and study methods have already been described.1

Data collection

Data on baseline demographic, clinical, and angiographic

Results

Eighty of the 1,054 (7.6%) patients enrolled in CABRI were excluded; 4 patients were lost to follow-up, and the other 76 exclusions were for incomplete baseline angiographic data and no procedure undertaken after randomization. The remaining 974 patients (92.4% of those enrolled in CABRI) formed the study population, all of whom had ≥1 year of follow-up available. Of these, 498 underwent PTCA and 476 CABG. Comparison of baseline variables between those in the present study population and those

Relation between revascularization mode and mortality

Logistic regression models based on baseline clinical variables adjusted for the level of CAD measured by the Duke score, or by the ≥50% score, indicate that having undergone a PTCA as the initial revascularization procedure is an independent predictor of mortality (i.e., OR 2.12 [p = 0.047] and OR 2.18 [p = 0.05], respectively). Without such adjustment, the initial revascularization with PTCA fails to reach statistical significance (RR 1.84, p = 0.092). In addition to the Duke score, age,

Acknowledgements

Most of the funding for the CABRI trial was at the expense of, and due to enthusiasm of, the individual participating centers whose imagination and perseverance made this trial possible.

References (26)

  • Coronary angioplasty versus coronary artery bypass surgerythe Randomised Intervention Treatment of Angina (RITA) trial

    Lancet

    (1993)
  • C.W Hamm et al.

    A randomised study of coronary angioplasty compared with bypass surgery in patients with symptomatic multivessel coronary disease (GABI trial)

    N Engl J Med

    (1994)
  • D Carrie et al.

    Five year outcome after coronary angioplasty versus bypass surgery in multivessel coronary artery diseaseresults from the French Monocentric Study

    Circulation

    (1997)

    • Cited by (0)

    ☆

    The CABRI Trial was sponsored by educational and research grants from CR Bard (USCI) Inc., New Jersey; The World Health Organization, Geneva, Switzerland; and The European Society of Cardiology, Sophia Antipolis, France.

    View full text
    Copyright © 2000 Excerpta Medica Inc. All rights reserved.