Improving mortality of coronary surgery

Editor—The findings of the study by Bridgewater et al of improving mortality of coronary surgery over the first four years of independent practice in 15 surgeons are relevant and clinically important.¹ But aren't the conclusions paradoxical?

The authors report that mortality in patients operated on by newly appointed consultant surgeons is similar to mortality in patients operated on by established consultants. Yet new consultants experience a halving in mortality after four years when they become established consultants. The reason is that Bridgewater et al used their peers as the benchmark to qualify the first statement, and their own results for the second.

This modest group of surgeons have underplayed their own good results that mortality to expected mortality of 0.36 compared with the additive EuroSCORE is impressive. Unfortunately, this affects the conclusions. The advantage that the authors used themselves as a benchmark for established consultants offers a more valid comparison. Case selection, operative performance, and management of postoperative complications will inevitably be more similar. If so, the conclusions would say that mortality in patients operated on by newly appointed consultants is twice that of established surgeons (the authors using themselves as internal controls).

Figure 1.

The real yardstick would be the next 15 newly appointed surgeons to the Northwest region comparing their results with established surgeons, as these new consultants would compare their initial results with a lower Northwest average (because of the contribution of this excellent newly appointed cohort). In this way a more accurate comparison would be obtained to determine if the results of newly appointed surgeons are on a par with their established peers.