Angina: CABG led to less reintervention than angioplasty, but mortality and MI were the same.

Clinical bottom line (level 1b)

  1. Patients with severe angina who had angioplasty compared with coronary artery bypass were less likely to die or have a Q-wave MI in hospital (NNT = 37 at unknown) .
  2. More required emergency CABG or angioplasty during their stay in hospital (NNH = 11 at unknown) .
  3. There was no difference in mortality or Q-wave MI at 5 years between the two procedures.
  4. More patients who had angioplasty required another revascularisation procedure within 5 years (NNH = 2 at unknown) .
  5. Fewer patients with diabetes who had CABG died than those who had angioplasty (NNT = 8 at unknown) .
The Bypass Angioplasty Revascularization Investigation (BARI) Investigators : New England Journal of Medicine 1996; 335 (4): 217-225
Expires July 2003

The study

Single-blinded concealed randomised trial with intention-to-treat
Setting: 16 centres in USA and 2 in Canada

1829 patients (aged mean 62 years, 73% male) angiographically documented multivessel coronary artery disease and either:
  • clinically severe angina
  • objective evidence of ischaemia
that required revascularisation and were suitable candidates for CABG or PTCA.
Note:
  • Intervention was performed within two weeks of randomisation.
  • Risk factor modification was emphasised to primary care physicians.


  • Control Group: (n = 914, 914 analysed): coronary artery bypass surgery
    Experimental Group: (n = 915, 915 analysed): coronary angioplasty

    98% followed for 4.5 years

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)
    ARR
    (95% CI)
    NNT
    (95% CI)
    death discharge unknown 12
    (1.3%)
    10
    (1.1%)
    17%
    (-92% to 64%)
    0.002%
    (-0.008% to 0.012%)
    460
    (NNT = 130 to infinity;
    NNH = 82 to infinity)
    Q-wave MI discharge unknown 41
    (4.5%)
    19
    (2.1%)
    54%
    (21% to 73%)
    0.024%
    (0.008% to 0.040%)
    42
    (25 to 130)
    death or Q-wave MI discharge unknown 52
    (5.7%)
    27
    (3.0%)
    48%
    (18% to 67%)
    0.027%
    (0.009% to 0.046%)
    37
    (22 to 110)
    emergency CABG or angioplasty discharge unknown 1
    (0.1%)
    57
    (9.2%)
    -8291%
    (-60000% to -1071%)
    -0.091%
    (-0.110% to 0.072%)
    -11
    (-14 to -9)
    death 5 years 382
    (40.7%)
    378
    (41.3%)
    -2%
    (-13% to 9%)
    -0.006%
    (-0.051% to 0.039%)
    -164
    (NNT = 20 to infinity;
    NNH = 26 to infinity)
    Q-wave MI 5 years 429
    (47.0%)
    428
    (46.8%)
    1%
    (-10% to 10%)
    0.003%
    (-0.043% to 0.048%)
    370
    (NNT = 23 to infinity;
    NNH = 21 to infinity)
    revascularisation 5 years 73
    (8.0%)
    499
    (54.5%)
    -583%
    (-758% to 444%)
    -0.465%
    (-0.502% to 0.429%)
    -2
    (-2 to -2)

    a priori subgroup analysis in diabetics
    Outcome Time to outcome CEREERRRR
    (95% CI)
    ARR
    (95% CI)
    NNT
    (95% CI)
    death discharge unknown 87
    (48.3%)
    104
    (60.1%)
    -24%
    (-51% to -3%)
    -11.8%
    (-22.1% to 1.49%)
    -8
    (-67 to -5)

  • No other subgroups had any significant differences.
  • Comments

    1. Increased number of deaths and Q-wave MI for CABG group may reflect the wait until operation can be performed.
    2. Not all cardiac events may be accounted for as non-Q-wave infarcts were not studied
    3. Good baseline study to compare results of future studies with the use of stens in angioplasty

    Citation

    1. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators , : Comparison of coronary bypass surgery with angioplasty in patients with multivessel disease. New England Journal of Medicine 1996; 335 (4): 217-225
    Contributor: Nick Shenker and Chris Ball, July 2000
    Reviewer: William Rhoton

    Clinical Question.
    Patient clinically severe angina
    Intervention or Exposure PTCA
    Comparison CABG
    Outcome death, revascularisation procedure