"Dad needs a bypass."
"Are you kidding?"
"No, really, all of the major blood vessels to his heart are critically blocked and the doctors think it's best to proceed with bypass soon."
"So what's the hang up?"
"It seems they found he has a 95% narrowing of his carotid artery on the right, too. They're worried he might have a stroke if the put him of the heart bypass machine."
There aren't too many more complicated issues for recommending a patient with severe coronary disease for bypass when there's a critical narrowing of a carotid artery. It is incredibly harrowing to fight the battle of coronary artery revascularization, only to lose the war when a patient wakes with an expressive aphasia (inability to speak) or the ability to move one side of their body after suffering a major stroke during bypass. This is not minor issue.
So today's article in Neurology with gushing claims from the lay press tries to shed a bit of light on the issue, claiming a 40% increased risk of stroke exists if a carotid endarterectomy is performed in conjunction with a heart bypass procedure, rather than as separate procedures.
But caution must be exercised when interpreting these researchers' findings. There is a clue to the problems with this trial: why was such a "cardiovascularly-related" article found in the journal Neurology? Could it be that the data are suffering from a homonymous hemianopsia?
In evaluating this work, the reader and lay press would be well-advised to review the methodology of this study. It used retrospective chart review of computer-coded data, albeit in large numbers of charts, in an attempt to glean a flicker of data with which to draw a glimmer of a trend - NOT a conclusion. To attempt to make any sweeping treatment recommendations (e.g. there is a higher risk of stroke with combined bypass and endarterecy) without knowing the severity of carotid narrowing, or even if stroke victims had both carotid arteries narrowed as opposed to one, quickly demonstrates the flaws in such a retrospective analysis. Teasing out the validity of data requires "retrospective" analysis that can be subject to bias as well - many of which cannot be anticipated by the reviewer - like coding bias - wich may have been performed to improve reimbursement by the hospital. Another bias might be changes in operators or surgical technique that occurred over the time period studied. Can the authors prove this did NOT happen with their retrospective evaluation? Of course not.
But the most damning of the findings of the study were the admission of the investigators themselves:
"The limitations to the use of administrative data sets include both inaccuracies and inadequacies of available data. Diagnostic coding errors are common, though improving over time. We have tried to eliminate as many miscoded cases as possible by narrowing the data set."Translation: "We know the data are poor and we fixed them a bit to clean them up and erased some data, but don't worry about that."
What can be said is that there were alot of computer-generated codes flying through a microprocessor and a statistics package that suggested a trend in increased risk might exist. To say much more with this study is meaningless.
A better source is the data from the 2004 ACC/AHA Guidelines for Coronary Artery Bypass Surgery:
Hemodynamically significant carotid stenoses are thought to be responsible for up to 30% of early postoperative strokes. The trend for coronary surgery to be performed in an increasingly elderly population and the increasing prevalence of carotid disease in this same group of patients underscore the importance of this issue. Perioperative stroke risk is thought to be <2% when carotid stenoses are <50%, 10% when stenoses are 50% to 80%, and 11% to 19% in patients with stenoses >80%. Patients with untreated, bilateral, high-grade stenoses and/or occlusions have a 20% chance of stroke. Carotid endarterectomy for patients with high-grade stenosis is generally done preceding or coincident with coronary bypass surgery and, with proper teamwork in high-volume centers, is associated with a low risk for both short- and long-term neurological sequelae. Carotid endarterectomy performed in this fashion carries a low mortality (3.5%) and reduces early postoperative stroke risk to <4%, with a concomitant 5-year freedom from stroke of 88% to 96%.'nough said.