I'm busy working on my blog posts. Watch this space!
Using Shared Decision Making to improve Erik's hospitalization: A blog series looking at the truth behind medical interventions, Part 1
August 7, 2016
As a corollary to Erik’s hospital experience as described in our JAMA article from July 2016, we will discuss four medical interventions that the medical team felt were vital for Erik while he was hospitalized, none of which were discussed with him, and all of which he ultimately rejected. These include his having bypass surgery for a single coronary artery that was blocked 80%; taking warfarin for a brief episode of atrial fibrillation; using high dose statins for an undefined period after his heart attack; and using beta blockers indefinitely after his heart attack. As noted in the story, Erick questioned the need for bypass and, without explanation; it was no longer mentioned by his doctors. He adamantly refused warfarin despite very strong opposition to that decision by his medical team. And he stopped his high dose statins (he continues taking modest doses) and beta blockers after his discharge due to life-altering side effects.
All medical decisions are wrapped in a veil of uncertainly, and none can be made in a vacuum. Each patient has unique medical circumstances, is on a unique set of medicines, has a unique degree of risk he/she is willing to tolerate, and has a unique reaction to the interventions. In Erik’s case, his care was based more on protocol-driven decision making rather than an individualized patient-centered approach. His unique medical conditions, reactions, and perceptions were never considered in the medical team’s assessment. He never was a meaningful participant in his own care. After our article’s publication, several doctors wrote in stating that Erik’s decision to eschew many interventions, and my (Andy) willingness to facilitate a more nihilistic approach, was irresponsible. In fact, some stated, we were denying Erik the most beneficial care available. Thus, we will review what we know about these four interventions and assess what in fact the risks and benefits of them might be to Erik. Ultimately though, in the end, it is his decision as to how much risk he is willing to accept, and how many interventions he feels are appropriate to his own situation. There is no right and wrong answer.
Single Vessel Bypass: We will start by addressing the initial declaration by Erik’s medical team that he would likely need bypass surgery of his 80% blocked LAD artery before leaving the hospital. Erik had his two blocked arteries stented when he presented with his heart attack, but the cardiologists were unable to open his LAD with a stent. The need for bypass of his LAD was brought up several times to Erik early in his stay, even to the point that he was told he would be meeting with a cardiac surgeon, but the more he talked to me about the need for bypass, and the more he questioned why it needed to be done, the more the issue faded from the radar, so that later in his hospital course it was not even mentioned. Why would his doctors strongly suggest the need for bypass in his situation, and what do the data tell us?
In opening a narrowed artery with a stent or bypass, an assumption is made that we will reduce a person’s chance of having a subsequent heart attack. There is a certain degree of logic to this line of reasoning: open the artery before it completely blocks. Unfortunately, the cause of heart attacks, and the reality of where they actually occur, makes this assumption dubious. In fact, there is little evidence that bypass or stenting helps avert a heart attack at all, and in fact it likely causes more harm than benefit. We discussed why this is the case in a recent blog (CLICK FOR BLOG)
Heart attacks occur when plaque within a heart vessel ruptures, instigating a cascade of reactions that lead the body to try to mend that tear with a scab that blocks the artery and causes the heart attack. Ruptures of plaque do not typically occur in blocked arteries. Some studies suggest that most heart attacks occur in blood vessels that have minimal blockage and would appear normal on a stress test or catheterization. As the BRCT at the top of this blog shows, a vast majority of people who have a heart attack would have a normal stress test immediately before the event (650 out of 1000 have a normal stress test) because they do not have blockages. Rather than blockage, the presence of plaque is a better predictor of heart attack risk, and plaque cannot be “fixed” with bypass. (CLICK HERE to see our blog on calcium scores to demonstrate a simple method to ascertain whether someone has a lot of plaque deposition) In Erik’s case, it is very possible that the blocked artery that did trigger his heart attack may have been relatively open before its plaque ruptured. Had Erik had a stress test before his heart attack, his blocked LAD may have been demonstrated, but is unclear if any other blood vessel was blocked at the time. Had his LAD been “fixed” with bypass, his heart attack would still have occurred, since it was not caused by a blocked LAD.
Studies suggest that in people who have had a heart attack, and/or who have plaque in their arteries, that taking a statin and an anti-platelet agent like aspirin can help prevent a second heart attack in as many as 10% of people (100/1000 people benefit). However, when studies measure the value of bypassing or stenting blocked arteries vs simply using aspirin and statins, the results show no improvement (0/1000 improve), while 20/1000 people have significant adverse consequences.
In assessing the evidence, Erik and I saw very quickly that the risk of his having a bypass of his LAD likely exceeded any perceivable benefit, especially since he had no symptoms of angina, no evidence of congestive heart failure, and he would be taking a statin and aspirin. Had this information been discussed with Erik during his hospitalization by his medical team, using simple to understand BRCTs and basic information about the pathogenesis of heart attacks (all of which could have been accomplished in 15 minutes), Erik could have made an informed decision rather than being stressed by having an intervention pushed on him without any explanation as to why he needed it and what the consequences might be if he did not agree to have it. As a frightened patient fully relying on his medical team to guide him, Erik was vulnerable to saying yes to anything his team suggested. Only by talking to me and looking at the data on his own was he able to ask the correct questions and arrive at a reasonable conclusion. If he were a passive patient, he may well have received a procedure he did not need and which could have caused him harm.
A New York Times article from August 2016 discussed the proliferation of high-cost low-yield medical procedures that have dubious benefit but which doctors continue to push and which insurance continues to pay for. While bypass was not one of the procedures specifically addressed by the article (it has been widely discussed by many other books and articles), the overall conclusions of the authors can easily apply to Erik’s situation. Most procedures like cardiac bypass are not regulated by the FDA, and doctors are paid well whether a procedure is known to be effective or is essentially useless. In Erik’s case, his taking inexpensive and effective medicines delivers far less reimbursement to the doctor than if he had bypass of his LAD. Many doctors truly believe that the procedure is more effective based on their own experiences because, as the Times article finds, doctors tend to only remember successful interventions. Finally, the article notes what we discussed in our prior blog on bypass: doctors tend not to discuss accurate data with their patients. Had Erik’s doctors used the BRCTs in our bypass/stent blog to initiate a meaningful discussion with Erik, then a much more fulfilling outcome would have been likely.
Ultimately a discussion of risks and benefits can decrease over-treatment and help patients reach conclusions that are best for them. As we will see with atrial fibrillation in the next blog, such a discussion can be achieved simply by using BRCTs.