Cost comparison of four revascularisation procedures for the treatment of multivessel coronary artery disease: a commentary

The main non-medical approaches to the treatment of symptomatic obstructive coronary artery disease include percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG). There are two main groups of coronary stents: bare-metal stents (BMS) and drug-eluting (DES). Coronary stents help to tack up the balloon-induced dissection that occurs during PCI and allow for better lesion expansion, thereby reducing periprocedural complications compared with balloon-only approaches. DES offer the same advantages but also have a polymer-coating embedded with medication designed to inhibit the degree of neointimal proliferation, which is the main cause of in-stent restenosis. CABG also offers two main approaches: standard ‘on-pump’ bypass and the more recent variant of ‘off-pump’ CABG. In the latter approach, the aorta does not need to be cross clamped and the patient is not placed on a cardiopulmonary bypass machine (pump) for their operation. Each of these four approaches has a different cost structure and varying adverse clinical event rates, both in the periprocedural period and in longer-term follow-up. There is a wealth of randomised clinical trial (RCT) data examining these various strategies in different populations; specifically, there is an abundance of data comparing BMS versus DES1, on-pump CABG versus BMS2–4 and on-pump versus off-pump CABG5,6. There are large, ongoing RCTs comparing CABG with DES7,8. Most of these trials have been undertaken in stable outpatients with symptomatic angina. Unfortunately, there is no one randomised study examining all four revascularisation approaches in the same population, and key assumptions are required to put together a dataset on which to construct an economic analysis.

In this issue (pages 119–134), we present an economic analysis of a four-way comparison of these different revascularisation strategies in stable, symptomatic patients with multivessel coronary artery disease. Since much of the data used to construct the base model was derived from RCTs with a follow-up period in the range of 1 year, a 1-year time horizon was evaluated. Given the lack of direct comparative RCT data amongst all four of these strategies, unreasonable clinical assumptions could clearly bias the outcome of the economic analysis. With a paucity of clear, unbiased RCT data extending beyond 1 year, this analysis was limited to a relatively short time period; important differences between these strategies extending beyond 1 year may emerge but are difficult to estimate. In addition, since the majority of the RCT data used to construct our database examined stable, relatively low-risk outpatients, one should not necessarily assume that the results can be easily generalised to higher risk patients such as those with recent acute coronary syndrome (ACS)9,10. There is currently a paucity of RCT data available, especially with CABG, examining these various revascularisation approaches in an ACS setting. Such higher-risk ACS patients, and the medications that they are prescribed, are associated with higher mortality rates, increased bleeding and other adverse outcomes with CABG compared with stable outpatients10. The optimal timing of inpatient CABG surgery has not been established and a cautious approach to higher-risk patients may be prudent9. Concerns also exist for PCI, including a higher rate of major bleeding, vascular complications and acute stent thrombosis. Thus, our model is best applied to a stable, symptomatic population with multivessel coronary disease.