Verified by 
Decision-Making
Can You Trust Your Doctor to Give You the Right Pill?
A common dilemma in medical decision making
Posted March 20, 2012
Susan Holmgren was resting in her hospital bed, a mere 72 hours removed from her first heart attack. She was out of the woods according to her doctors, with her heart having only sustained minor damage. But the next one-the next heart attack-could be worse. Almost 75-years-old now, Holmgren needed to do everything in her power to make sure that her first heart attack would be her last. So Holmgren decided to quit smoking. She committed herself to losing weight and exercising. And she agreed to take all the new pills her doctors prescribed.
But was one of those pills a mistake? And how should she, or her doctor, have done things differently?
Susan Holmgren is a fictional patient. But I made up her case to illustrate a common challenge in medical decision making-how do we take evidence from clearly designed research trials and translate them into the messy real world?
In January of 2012, the New England Journal of Medicine published a landmark trial testing whether a new medication, Xarelto, reduces heart attacks in people who have already developed coronary artery disease. The research team enrolled over 15,000 patients into the trial-a Herculean task that no doubt cost the company hundreds of millions of dollars-and randomly assigned patients to receive either the standard of care in heart attack prevention (pills like aspirin and beta blockers) or to receive Xarelto in addition to the standard of care. The results were impressive. Over the two-and-a-half years of the study, low dose Xarelto cut deaths from 4.5% to 2.9%. That means for every 60 people who receive Xarelto, one had his or her life saved by the drug.
Xarelto isn't without risks. Like many heart attack preventing drugs, it "thins" the blood and thus can cause bleeding. But it is not the side effects that cause me to ask whether patients like Susan Holmgren should take Xarelto when their doctors prescribe it. Instead it's another problem: most of the 15,000 people in that clinical trial aren't as sick as Holmgren was when she started taking that pill.
The Xarelto trial was a paradigm of good medical science. The researchers selected patients carefully, including those with a recent history of stomach ulcers for example, out of fear that those people would bleed again with the new pill. They also excluded people with any history of kidney disease, probably out of concern that the dosage of the medicine wouldn't be as stable in such people.
But Holmgren had a touch of kidney failure, and she was older than most of the patients in the study. And she was, . . . well, . . . a she. The vast majority of the patients in the Xarelto trial were men, a fact that makes it hard to know whether the results of the trial really applied to her.
As I imagine this story, Holmgren's doctor clearly thought the benefits of the medication outweighed the risks. She was a "young seventy-five." Her kidney failure was mild, really mild. And the physician couldn't bear the thought of Holmgren experiencing another heart attack when a great drug was available that would potentially reduce that risk so dramatically.
Was that the right decision? Well let's consider the doctor's alternatives. First, he can decide never to prescribe treatments until there is solid scientific evidence that they benefit patients similar to the patient at hand. By that logic, he would not offer Holmgren Xarelto. By that same logic, he would have a whole lot of tough decisions to make about other patients, because most careful medical trials enroll patients who never quite fit the description of the patient sitting in front of the doctor in the exam room. It would be hard to give many treatments to women, because they are so frequently underrepresented in clinical trials. It would be even more difficult to give any treatments to African Americans, who made up less than 1% of the patients in the Xarelto trial, an all too frequent phenomenon in medical research. Should all these patients be denied the potential benefits of this treatment? After all, there is little chance that researchers will conduct a follow-up study of 15,000 African American women. Clinical trials are insanely expensive. Consequently, scientific evidence will never be easily applied to all patients.
This doctor has a second alternative: simply to make a judgment call of whether this drug is likely to benefit this patient, given the similarities and differences between her and the 15,000 patients in the trial. This is not an easy judgment to make. But it's what doctors are trained to do.
A third option is to wait for the experts to chime in, to develop guidelines that determine which patient should be given which drugs. Ideally these guidelines won't be biased by, say, industry influence. Sadly, most such guidelines are usually created by clinical trial experts, most of whom have deep and complicated ties to industry.
Fourth, the doctor could discuss the pros and cons of Xarelto with Holmgren and let her decide whether to take the pill. In a world where we tout the importance of shared decision making, this option seems like it ought to be the best. If the right treatment for Holmgren isn't straightforward, and there are risks and benefits of Xarelto that Holmgren needs to know about, maybe she should be part of this decision. But it's not easy to involve patients in these kinds of decisions. I read through the New England Journal article and tried to imagine what I would say to a patient like Holmgren. I couldn't wrap my mind around the data well enough to describe it to her in a pithy and informative way.
Finally, the doctor could employ a softer form of informed consent in such cases. Doctors need to make difficult judgments about which patients to prescribe which therapies. As the judgment becomes more difficult, they can explain to patients why they're prescribing the drug, and why there is clinical uncertainty about what's best in their case. Maybe this doctor could have said something to Holmgren along the lines of: "In addition to the aspirin and beta blockers, I am giving you a new pill called Xarelto. A study in Europe showed it helps patients to avoid heart attacks, but the patients in the study were mainly white men who are younger than you. My guess is that you'll benefit from this pill. But it does carry the risk of bleeding. And if you're nervous about that, I totally respect it if you want to hold off on taking this medication."
I prefer this last option. But I cannot say I have consistently been so forthcoming with all my patients. It is too hard to keep track of all the trials and all the uncertainties.
No one ever said medical decision making was easy.
