Chapter 12.4: Composite End Points

CLINICAL SCENARIO

You are an internist seeing a 65-year-old man with stress test–documented angina who—despite taking carefully titrated β-blockers, nitrates, aspirin, an angiotensin-converting enzyme (ACE) inhibitor, and a statin—is substantially restricted in his activities. The patient undergoes coronary angiography, which reveals 3-vessel severe coronary disease. You suggest to him the possibility of surgical revascularization with coronary artery bypass grafting (CABG). The patient expresses reluctance to undergo such an invasive procedure, and he asks if there is a less aggressive approach that might be almost as effective. You consider the possibility of a percutaneous coronary intervention (PCI) as an alternative.

You wonder what recent evidence might bear on the patient's dilemma. You ask the patient to join you in front of your computer, and you go to the online version of ACP Journal Club, which you can access through your library subscription (http://acpjc.acponline.org/gsa-search). To guide your search, you jot down your question in PICO format: In patients with 3-vessel coronary artery disease, what is the impact of PCI vs CABG on angina, major cardiovascular event, and overall mortality? (See Chapter 4, What Is the Question?) Because ACP Journal Club selects only a small subset of clinically relevant studies, you decide to start with a broad search (see Chapter 5, Finding Current Best Evidence). You therefore enter search terms describing only the patient population: those with multivessel coronary artery disease. The search yields 16 citations, the second of which is a randomized clinical trial (RCT) of PCI vs CABG called Synergy between PCI with Taxus and Cardiac Surgery (SYNTAX).¹ You tell the patient you will review this study carefully and discuss the results with him in a week.

In the SYNTAX study, you find that 1800 patients with 3-vessel or left main coronary artery disease were randomized to undergo CABG or PCI. The study found a significantly lower rate of the composite end point—death from any cause, stroke, myocardial infarction (MI), or subsequent revascularization—in the CABG arm (12.4%) than in the PCI group (17.8%) (relative risk [RR], 0.69; 95% confidence interval [CI], 0.55-0.87; P = .002). The authors concluded that CABG should remain the standard care for patients with severe coronary artery disease.

How should you interpret these results to best inform your patient's decision? Should you assume that the effect of treatment on the composite end point accurately captures the effect on its components (death, stroke, MI, and subsequent revascularization)? Or, rather, should you look more carefully at each component and draw the individual effects to your patient's attention?