Coronary artery bypass graft surgery trends and outcomes in the UK: established excellence or still room for improvement?

In the current issue of the journal, Ohri et al. [1] describe trends and outcomes for coronary artery bypass grafting (CABG) in the UK over a 15-year period from 2002 to 2016.

The authors should be sincerely congratulated on this highly relevant and relatively up to date Society of Cardiothoracic Surgery-commissioned review. The Society of Cardiothoracic Surgery was one of the first cardiac surgery societies in the world to collect national data for surgical outcomes starting in 1977. In the 1990s, increased media, political and public expectations demanded that surgical outcomes be published, initially with hospital specific data in 2001 and then individual surgeon data in 2004.

There are several key strengths to the report. First, it is particularly comprehensive and covers the vast majority of CABG operations conducted in the UK and Ireland over the relevant 15-year period. Second, in addition to identifying general trends regarding isolated or combined CABG operations, it additionally reports more specific trends in certain patient subgroups.

The most dramatic observation is a 38% reduction in the annual volume of CABG operations from around 24 000 to 15 000 cases per year but with relatively stable numbers over the last 4 years of the study. Similar reductions in CABG volume have been reported from many other parts of the developed world including Europe and North America.

The key findings in terms of risk adjusted mortality are especially encouraging and compare very favourably with other international databases. For elective CABG, this decreased from 1.5% to 0.6%, for urgent CABG from 3.1% to 1.3% and for emergency CABG (around 2% of all CABG) from 11.7% to 6.6%. For combined CABG and valve operations (18% of all CABG procedures) mortality decreased from 7.9% to 4% while for re-do CABG (only 0.2% of the whole CABG population) mortality stayed at over 7%. Off-pump CABG decreased from a peak of around 20% in 2010 to now just over 10%.

However, the report also presents a number of less comfortable issues. First, the average number of CABG cases per cardiac surgery unit declined from a mean of 720 to449 cases and for individual surgeon from 124 to 74 CABG per year. And while there is no agreed absolute specified number of cases to either achieve or maintain competency for a particular operation, a general volume–outcome relationship is widely accepted and especially for more technically demanding operations where the margins for error are considerably smaller. A 1-mm error in a valve operation is unlikely to have a significant clinical impact but may have dire consequences for a distal anastomosis in CABG.

The second is that CABG results and implied quality are almost exclusively focused on short-term outcomes. While these are unquestionably important data to document, this approach fails to capture that another fundamental essence of CABG is also to provide long-term benefits of revascularization. Indeed, in the Arterial Revascularization Trial of single versus bilateral internal thoracic artery grafts, around 80% of patients were alive at 10 years, testifying to the durability of CABG [2]. Consequently, a relentless focus on exclusively short-term outcomes may be both a strategic error and, especially over the longer term, a false economy. Such an approach does not encourage interventions that may potentially improve long-term outcomes of CABG including the use of more than a single arterial graft.

Disappointingly over the period of this study, the use of more than a single arterial graft for CABG fell from 18% to 10%. This is particularly regrettable given the well-documented significantly superior long-term angiographic patency of arterial grafts. A substantial body of evidence from trials and observational studies, extending over decades, also documents that multiple arterial grafts can lead to superior long-term clinical outcomes [3, 4]. While the Arterial Revascularization Trial itself was overall neutral in terms of survival, there was substantially better long-term survival in those patients who actually received more than a single arterial graft [2]. And is there any surgeon or cardiologist, who would argue biologically or mechanistically that having more patent grafts on the heart at 10 or 20 years is not advantageous to the patient? Indeed, the current report also notes that over the period of the study the ratio of Percutaneous Coronary Intervention (PCI) to CABG increased from 1.3 to 1 to almost 7:1. Would the same have been true if most potential CABG patients would receive at least 2 arterial grafts?

A third issue is that the study does not address possible risk averse behaviour in a potentially punitive outcome system that both punish the individual surgeon while hiding wider failings in the delivery of safe surgical services [5]. And, paradoxically, while there is a strong perception that cardiac surgery patients today are older and sicker with more co-morbidities, the mean logistic EuroSCORE remained unchanged, respectively, in elective (3.1) and urgent (5.9) patients. So, while we know the outcomes for patients who underwent CABG, we do not know if patients with a potential beneficial risk/benefit ratio over the longer term were denied CABG because of a perceived increased short-term risk of surgery (and on which the surgeon is solely judged). There is some reassurance in a three-fold increase in the number of octogenarians undergoing CABG, to 7% of the CABG population, and that is expected to continue to increase. However, a robust 80-year old may be a better surgical prospect than a sick and frail 60-year old.

Finally, and understandably, the authors do not address the potential role for super-specialization, for at least some CABG surgeons [6]. While the ‘standard’ operation of an internal thoracic archery and 2 or 3 supplemental vein graphs performed ‘on-pump’ is an excellent operation for many patients, it is not ideal for all. There is strong evidence that off-pump CABG using an an-aortic (‘No Touch’) technique significantly reduces mortality and stroke and several other aspects of major morbidity [7]. While not every CABG surgeon needs to be skilled with this approach, every unit offering CABG should have at least some surgeons with appropriate expertise in this technique. The traditional and predictable refrain of the surgeon on the way to the operating room saying that ‘it’s just another CABG’, and implying that it is the same operation for every patient, is no longer consistent with the best service for individual patients or our specialty.

REFERENCES