Applicability of the COMPASS Trial to General Practice

Study Questions:

What proportion of patients in the REACH (Reduction of Atherothrombosis for Continued Health) registry would be eligible for the COMPASS (Cardiovascular Outcomes for People Using Anticoagulation Strategies) trial?


The COMPASS trial evaluated the effectiveness of aspirin 100 mg daily, aspirin 100 g daily plus rivaroxaban 2.5 mg twice daily, or rivaroxaban 5 mg twice daily in patients with stable coronary artery disease (CAD) or peripheral vascular disease (PAD). The primary outcome was a composite of cardiovascular (CV) death, stroke, or myocardial infarction (MI). The trial enrolled patients from March 2013 through May 2016. The REACH registry is an international, prospective observational registry that includes patients with atherosclerotic disease (CAD, PAD, or cerebrovascular disease [CVD]), or three atherosclerotic risk factors. This study includes patients enrolled in REACH from December 2003 until June 2004. Outcomes for patients in the REACH registry were compared with patients in the aspirin only arm of the COMPASS trial. In order to better match the profile of the patients in the COMPASS trial, patients enrolled in REACH with CVD or atherosclerotic risk factors, patients who did not meet inclusion criteria for COMPASS, as well as those whose eligibility for COMPASS could not be determined due to incomplete or missing data were excluded.


There were 65,531 patients in the REACH registry and 16,875 were COMPASS-eligible and included in this study. After patients were excluded because of missing data (12,606) or atherosclerotic risk factors only or CVD (21,052), 31,873 patients were COMPASS-evaluable. There were 5,480 REACH patients who did not meet the inclusion criteria for COMPASS and an additional 9,518 patients were excluded due to COMPASS exclusion criteria. Therefore, in the REACH registry, 52.9% of the COMPASS-evaluable patients were COMPASS-eligible. When compared with the aspirin only arm of COMPASS, patients in the REACH registry were older, more likely to be female, and less likely to be taking evidence-based secondary prevention medication and had a lower risk of bleeding and ischemic events.

The COMPASS-eligible patients in REACH were more likely to experience the primary outcome of CV death, stroke, or MI when compared with the COMPASS patients in the aspirin only arm (event rate per 100 patients/year: 4.2 vs. 2.9; p < 0.0001). All-cause mortality was also higher in REACH (3.2 vs. 2.2 per 100 patients/year; p < 0.001). CV death and heart failure hospitalizations were also more common in REACH patients when compared to the aspirin group in COMPASS. Due to different definitions, bleeding rates could not be compared.


About 53% of patients enrolled in the REACH registry would have been eligible for the COMPASS trial, suggesting that COMPASS is applicable to general practice, outside a clinical trial.


There has been concern that clinical trials may not be applicable to general practice because they enroll a highly selected cohort of patients. The authors used the REACH registry, which better reflects the patient population seen in typical practice, to see if these patients would be eligible for the COMPASS trial. The finding that slightly more than half of the COMPASS-evaluable patients in REACH were COMPASS-eligible suggests that the COMPASS findings could apply to general practice.

The differences in outcomes between the COMPASS and REACH patients likely reflect the higher use of evidence-based secondary prevention treatments in COMPASS. This difference in use of evidence-based medicine is best explained by the REACH patients being enrolled more than 10 years ago, when the data regarding the use of these agents were less robust. However, a sensitivity analysis done by the authors suggests that the difference in use of evidence-based secondary prevention medication is not the only factor. An alternative contributing factor is likely related to the different characteristics of patients enrolled in a clinical trial when compared to routine practice. A limitation of this work is that some clinical characteristics were defined differently in the two studies, meaning the comparisons should be interpreted cautiously.

While there are differences between the COMPASS trial and REACH registry, a substantial portion of patients in REACH would be eligible for COMPASS, suggesting that this trial is applicable to clinical practice.

Clinical Topics: Anticoagulation Management, Heart Failure and Cardiomyopathies, Prevention, Vascular Medicine, Atherosclerotic Disease (CAD/PAD), Acute Heart Failure

Keywords: Anticoagulants, Aspirin, Atherosclerosis, Cerebrovascular Disorders, Coronary Artery Disease, Evidence-Based Medicine, Fibrinogen, General Practice, Heart Failure, Hemorrhage, Myocardial Infarction, Outcome Assessment (Health Care), Peripheral Vascular Diseases, Primary Prevention, Risk Factors, Secondary Prevention, Stroke, Vascular Diseases

< Back to Listings