September 5th, 2012
Databases and Decisions
Harlan M. Krumholz, MD, SM
The decision about whether to undergo elective ascending aorta and/or arch replacement is often challenging. For some patients the indication may be obvious, but many patients need to weigh the potential morbidity and even mortality associated with the surgery against the chances that they will experience an acute, life-threatening event. Patients need good information about the risks and the benefits.
The current issue of the Journal of the American College of Cardiology has a report from the Society of Thoracic Surgeons that reports the outcomes of 27,202 patients who underwent elective ascending aorta and/or arch replacement surgery. Such studies of can provide good information about the experience of patients who do choose surgery even as they cannot illuminate the benefits.
The study reported a mortality rate of 3.4%, which the authors describe as excellent. But what we are not told is how the rate varies across centers: Did all the hospitals have roughly similar rates or did mortality vary widely by institution? Ultimately, whether 3.4% mortality is “excellent” may depend on whether you are the surgeon or the patient. We are not given information about whether it is as low as it could be. Finally, this study shows that major morbidity is much more common than mortality: stroke or coma occurred in 3.2% of patients, renal failure in 4.4%, pneumonia in 4.1%, reoperation for bleeding in 5.7%, and prolonged ventilation in 16.2%. The risk seems considerable for an elective procedure.
This study does provide critical information about contemporary practice. I wonder how this information will be incorporated into shared decision-making practices across the Society for Thoracic Surgeons sites. And will the risk estimates be tailored for patients and for the site?
2 Responses to “Databases and Decisions”
NEJM Journal Watch — Recent Cardiology Articles- Sex-Based Disparity in CABG Outcomes: Are We Moving Toward Parity?
- Can Genetic and Social Risk Scores Improve Cardiac Risk Prediction?
- Extensive Ablation vs. Pulmonary Vein Isolation in Persistent Atrial Fibrillation
- Can a Novel Stent Design Improve Long-Term Outcomes of Percutaneous Coronary Intervention?
- Are Angiographically-Derived Coronary Physiology Measurements Ready for Use in the Cath Lab?
I agree totally. It is the patients who are the beneficiaries or the victims of this surgical intervention, and there should be patient input on what an “acceptable” mortality rate is. And I wonder what the patients were told about the % risk of morbidity and mortality before they underwent the procedure. I would have more respect for the claimed result if those numbers were made available. And what % of patients who had the problem and refused the surgery went on to suffer death from this condition when untreated? That is a very important number as well.
I suppose the presence of millions of cases in the STS database compels one to extract data and try to distill a statement of truth. I actually believe that there is some good data mixed in with the unfortunate use of the adjective “excellent” to describe mortality – as if any mortality could be excellent. And yes it is the patients who benefit or lose. The value in this data analysis lies not in the reported mortality rate, but in the comparative between elective and non-elective proximal aortic surgery (3.4% vs 15.4%, respectively). It may be common sense to say that non-elective cases result in higher mortality, but, to my knowledge, this is the largest group of examined patient records and probably represents truth. Unfortunately the authors allude to the fact that the elective cases are most likely not the same as the other cases. As anyone who has suffered through the usual 2 am acute aortic dissection can attest, the operative field is a battleground filled with weapons of mass destruction, with the surgeon doing his/her best to sew two pieces of wet tissue paper together and then maintain hemostasis under pressure. This is certainly not the case of elective repair of an ascending aortic aneurysm, with or without a root replacement.
What the authors should have discussed more cogently is the knowledge that progression from elective to emergent, that is, from dilation to rupture or dissection, is not a completely predictable event. The best information about this probably comes from the Yale group who looked at 1600 patients followed there, having an average annual rate of dilation of 0.1 cm. They found that inflection points for bad events like rupture and dissection were about 6 cm in the proximal aorta, that, upon reaching this point, the risk of rupture or dissection leading to a more than 10% mortality rate was greater than the risk of death in the elective setting (2.5% at Yale).
Thus, the editors of the current study should be chastised for allowing qualitative descriptions of mortality to make it into print. What is indeed excellent is the knowledge that we have a better chance of saving lives by identification of a group at risk, and that the survival advantage goes up by a factor of eight. And that this is the case even when the average number of proximal aortic procedures performed per hospital is 12 patients per year.