Guidelines Should Change When the Data Say So
For decades people with even low risk for having a heart attack or stroke have been taking low-dose aspirin (usually 81 mg) daily to prevent a first heart attack or stroke. Perhaps because it’s called “baby” aspirin, it sounds so benign. What could be risky about something with the word “baby” in it?
In an excellent article in the New York Times last October health columnist Tara Parker-Pope wrote “…it came as a shock to many this month when an influential expert panel, the U.S. Preventive Services Task Force, seemed to reverse decades of medical practice, announcing that daily low-dose aspirin should no longer be automatically recommended in middle age to prevent heart attack.” To some, Parker-Pope noted, that “shock” translated into the belief that medical experts had “flip-flopped” on the issue. Some people, she implied in the article, will lose faith in medical recommendations because of what seem like abrupt changes in advice.
But Parker-Pope also does an excellent job of tracing the evolution of recommendations to take (or not to take) aspirin to prevent cardiovascular (e.g., heart attacks) and cerebrovascular (e.g., strokes) events and convinces us that this is not random flip-flopping but rather the orderly process of science.
Why Would Aspirin Work?
To start, it is logical to ask the question of whether aspirin would be expected to have any effect in reducing the risk a person has to have a heart attack or stroke. We know that many heart attacks and strokes are caused by blood clots forming in arteries that carry blood to the heart and to the brain. Some of these blood clots are thrombotic, arising in those arteries themselves, and others are embolic, arising somewhere else like a deep leg vein or the surface of a heart valve, and travelling through the circulation until they get stuck in a small artery. Either way, those blood clots reduce the ability of blood to flow through the artery where they lodge, decreasing the amount of oxygen that can be delivered to the heart or brain tissue, and thus causing cells to die. Because strokes and heart attacks are high on the list of the most common causes of death, it makes sense to try to prevent blood clots from forming.
The physiology of blood clotting is complex and involves an array of elements, one of which is a cell that circulates in the bloodstream called the platelet. One of the many actions of aspirin is to reduce platelet activity, thereby making the formation of abnormal blood clots less likely. So aspirin has been seen as one way of reducing the potential for abnormal clotting to occur.
At First It Seemed To Work
The original study, as Parker-Pope points out, that seemed to suggest a benefit for aspirin in preventing first heart attacks—in men—was published in the New England Journal of Medicine in 1988. The study involved more than 22,000 male physicians who took one 325 mg tablet of buffered aspirin every other day or a placebo. Compared to placebo, participants in the aspirin group had a nearly 50% reduction in both fatal and non-fatal heart attacks. The effect was so large that the study was stopped by the study’s independent data monitoring board and the results published.
There was also a slight increase in strokes in the aspirin group compared to the placebo group in that study. Experts attributed this to what are called hemorrhagic strokes—strokes that occur when a blood vessel bursts and blood leaks into surrounding brain tissue. Because aspirin decreases the blood’s ability to clot by interfering with platelets’ ability to form clumps, these strokes were probably an adverse side effect of taking aspirin. At the time, it seemed clear that the benefits of aspirin outweighed the risks, and many people began taking aspirin in various preparations and at different doses for primary prevention—that is prevention of first—heart attacks. Aspirin also came to be recommended by some guidelines for secondary prevention of heart attacks and strokes, that is to prevent them in people who already had serious cardiovascular or cerebrovascular disease or previous heart attacks and strokes.
In its 2016 guidelines, the U.S. Preventive Services Task Force (USPSTF), an independent body of experts that reviews literature and data to make recommendations about how to prevent diseases from occurring, made the following statement about people between the ages of 50 and 59:
The USPSTF recommends initiating low-dose aspirin use for the primary prevention of cardiovascular disease (CVD) and colorectal cancer (CRC) in adults aged 50 to 59 years who have a 10% or greater 10-year CVD risk, are not at increased risk for bleeding, have a life expectancy of at least 10 years, and are willing to take low-dose aspirin daily for at least 10 years.
But in its 2021 guidelines the USPSTF says the following about people between the ages of 40 and 59:
The decision to initiate low-dose aspirin use for the primary prevention of CVD in adults ages 40 to 59 years who have a 10% or greater 10-year CVD risk should be an individual one. Evidence indicates that the net benefit of aspirin use in this group is small. Persons who are not at increased risk for bleeding and are willing to take low-dose aspirin daily are more likely to benefit.
Now, that is a big change, from a blanket recommendation that people with a relatively low risk of heart attack start taking low dose aspirin on a daily basis to making the decision “an individual one” and saying the net benefit “is small.” Parker-Pope nicely points to three studies published in 2018, that either failed to show any benefit of aspirin in preventing heart attacks or found that the benefit didn’t outweigh the risk of bleeding as an adverse side effect of aspirin. Those new studies influenced the USPSTF to propose altering its guidelines for aspirin use, which are still being worked on by the agency.
What Accounts for the Difference in Findings?
What changed between the 1988 study that seemed to clearly indicate a benefit for aspirin to 2018, when equally well-designed studies did not? While it is possible that the 1988 results were a fluke, that seems unlikely given the magnitude of the findings. But remember that the 1988 study only included male physicians and they are hardly a representative group of the entire population. When women and people who aren’t doctors are included in more recent studies, things might be expected to change.
Parker-Pope makes an interesting speculation about what else might have changed since 1988—our general health. She writes “Fewer people smoke, and doctors have better treatments to control diabetes, high blood pressure and cholesterol, issues that all affect risk for heart attack and stroke. Aspirin still works to protect the heart, but doctors say the benefits aren’t as pronounced now that other more effective treatments have emerged. As a result, the risks of aspirin, including gastrointestinal bleeding and brain hemorrhage, are of greater concern, though they remain low.”
It is plausible, as Parker-Pope is suggesting, that 20 years ago the benefit of daily aspirin in preventing fatal heart attacks was larger than the risk from daily aspirin of serious bleeding, like hemorrhagic strokes. But as the risk to the population of having a fatal heart attack decreased because of things like less cigarette smoking and better control of high blood pressure, the added benefit of daily aspirin becomes no longer large enough to outweigh the risk of bleeding. This could account for the difference in findings between 1988 and 2018.
There’s a lot more to these recommendations. We’ve only touched here on primary prevention for people who don’t have much risk for heart disease. Other recommendations apply to primary prevention in people who do have higher risk and to secondary prevention. One expert stated emphatically that “The easiest patient group to address is adults of any age who have a history of heart attack, stroke, or revascularization [e.g., having had a coronary artery stent placed] and are taking aspirin for secondary prevention. They should continue taking aspirin; the new recommendations don’t apply to them.” There are now more drugs that inhibit clotting than we had in 1988, so aspirin may not be the right choice for many or even most patients who should be on some form of anticoagulation therapy to prevent heart attacks and stroke.
The important point here is to notice that as new science gives us new data, guidelines are going to change. That is not flip-flopping, it’s science. It is of course critical that physicians and other healthcare providers are conversant with the data and the latest iterations of treatment guidelines. People should not despair or be frustrated when recommendations change if the changes are made on the basis of new and emerging science. The complete story about aspirin use to prevent cardiovascular and cerebrovascular events is probably not written yet; there will certainly be more studies reporting more data, some of which may complicate the picture. We need to be sure that the science is allowed to flow unhindered and that the information about the results of that science are carefully interpreted by experts and made widely available to the public.
Several things can be done to help people accept changes in scientific consensus. We can continue to hammer at journalists and their editors to be careful how they present new findings, refraining from calling everything a “breakthrough” and acknowledging in their stories that almost every new finding raises important questions that will be researched further. Then, journalists and editors could report more often on ongoing research that hasn’t yet reached the level of changing guidelines for care so that people can see how the process of scientific advances evolves over time.
More fundamentally, we need to educate people from elementary school on how science really works. Once science was taught to children and adolescents as a fixed group of facts, creating the impression that everything was set in stone and making it hard for anyone to accept change. Even with the evolution of more hands-on learning, in which students are encouraged to do “experiments” and to solve problems, there has been an emphasis on “getting the right answer.” “Experiments” in science class are often more like recipes and are supposed to lead to a single, correct final result. What we need is to show students is that experiments rarely lead to a set of totally “correct” or expected results, that many experiments either produce unexpected and difficult to interpret results or fail altogether to produce a significant finding. From there, scientists keep designing more experiments, trying to work out what went wrong the first time(s) until they get something interesting. We need, therefore, to help people cope with uncertainty and change in science.
criticascience 