The effect of adjusting for baseline risk factors and post revascularisation coronary disease on comparisons between coronary angioplasty and bypass surgery

Background: In CABRI at 1 year PTCA was associated with greater repeat revascularisation and angina (but not myocardial infarction or death). We determined whether adjusting for baseline risk factors and post revascularisation coronary disease offsets this disadvantage of PTCA. Methods: In the CABRI population the crude association of revascularisation mode (i.e. PTCA or CABG) with four clinical outcome (i.e. mortality, myocardial infarction, repeat revascularisation and angina) was adjusted for the baseline risk factors using a logistic regression model for each clinical outcome. A number of measures of angiographic coronary disease were used to assess post revascularisation coronary disease. One at a time, each of these measures was added to each of the four outcome models, to adjust for post revascularisation coronary disease. Results: Comparing adjusted and crude unadjusted association of PTCA with repeat revascularisation there was an increase from 12.8 (P<0.0005) (crude relative risk) to 16.7 (P<0.0005) (adjusted odds ratio), with angina, from 1.89 (P=0.001) to 1.98 (P<0.0019), and with mortality from 1.84 (P=0.092) to 2.15 (P=0.060). PTCA was not significantly associated with myocardial infarction, either crudely or after adjustment. Conclusion: Adjusting for baseline risk factors and post revascularisation coronary disease tended to strengthen rather than weaken associations between PTCA and 1 year mortality, repeat revascularisation and angina at 1 year.