Despite vast differences among patients, major clinical guidelines assume that everyone should get the same type of care so as to ensure the best possible outcome for all. This egalitarian approach to patients is considered an ethical touchstone of modern medicine. But all patients are not created equal. When it comes to following a doctor’s advice, taking pills on a regular schedule, or otherwise following through on a given course of treatment, some people perform better than others. And that means we might be better off tailoring our guidelines to match the skills and proclivities of each individual patient.
Consider the case of asthma. The standard treatment is two-pronged: first, an immediate relief inhaler to stop wheezing, and second, “controller” medications that reduce lung inflammation and ward off future flare-ups. As with many preventive drugs, the controllers must be taken every day to be effective—even when patients feel fine. We also know that patients should try three or more combinations of drugs before finding the one that works best for them, in what is an expensive and labor-intensive process. If every patient dutifully followed the instructions of his or her doctor, this would all be worth it: Asthma-sufferers would always end up with the best treatments available, no matter how complex. But that’s not how things play out in the real world.
As reported a few months ago in the New England Journal of Medicine, researchers in the United Kingdom tried something different with controllers. Instead of comparing one against another in a well-orchestrated clinical trial, in which patients are seen more often than normal, get reminders to take their meds, and receive special incentives for cooperation, the researchers tried to “preserve the ecology of clinical care.” In other words, they ran the study though primary care doctors instead of a fancy academic center, and didn’t give out any patient bribes. After two years of study, they found it didn’t really matter what controllers the patients got, as most of them never bothered to take the medicines anyway. (Keep in mind this occurred in a single-payer system, so patients’ incomes were irrelevant.)
The findings from this so-called “pragmatic trial” may change the way British doctors prescribe drugs for the prevention of asthma flares. Patients may be told that it doesn’t really matter what controllers you take, although intensive clinical trials suggest otherwise. In effect, the recommendations will be dumbed down, because some (or even most) patients simply aren’t very good at following directions. Even savvy patients—the ones who are perfectly able to handle the more complex and better treatments—would be treated the same as everyone else.
Should patients be separated by ability groups, as some students are in schools? Tracking students is controversial. Generally, those focused on the high-achieving students prefer such grouping, on the theory that their potential is wasted in typical classrooms. Advocates for lower-performing students, though, oppose it, saying that tracking dooms them with low expectations. How does this debate play out in medicine?
Consider another example, involving adults with coronary artery narrowing, which is often treated by the insertion of a metal stent to prop open the vessel. Several studies show that so-called “drug eluting” stents work better than conventional bare metal stents for these patients. But there’s a catch. If patients don’t dutifully take anti-clotting drugs every day for at least a year, the drug-eluting stents are more likely to clog suddenly and kill them. So if you wanted to improve care for everyone, you’d give drug-eluting stents to the A+ patients—those who are most likely to take their medications as instructed—and give the cheaper, safer (but less effective) metal stents to the remedial or unresponsive patients. But no clinical guideline makes this point.
These dilemmas occur throughout medicine. The only really effective therapy for severe cystic acne is isotretinoin, better known as Accutane, which causes great harm to developing fetuses. Before any woman can get it, she must commit to being on 2 separate forms of birth control and getting monthly pregnancy tests. That is to say, the guidelines assume that every female patient is at the back of the class: Even responsible young adults who either abstain from sex or are fastidious about condom use while taking Accutane are often forced into taking birth control pills.
Many insurance companies, in general, lower the bar and set guidelines according to the least common denominator. Consider insulin pumps, which are fancy devices that more accurately infuse insulin and prevent blood-sugar swings in highly compliant diabetics, but can harm careless patients who operate them wrongly. Insurers make little effort to connect cooperative patients with this superior, but more complex, treatment.
Like patients, doctors also have different ability levels. The American Academy of Pediatrics discourages the use of antibiotics for toddlers’ ear infections, on the basis of clinical data showing the drugs don’t work. But as the New England Journal of Medicine recently reported, antibiotics really do help when well-trained doctors are involved. The key, an accompanying editorial dryly pointed out, is “the accuracy of the diagnosis,” since many doctors mistake simple colds for ear infections. In other words, doctors who can’t diagnose ear infections reliably shouldn’t prescribe antibiotics, but those who know their stuff should. Again, no guideline makes this distinction, and the academy caters to the most mediocre doctors with its broad warnings.
The trade-offs between standardization and personalization lie at the core of our national health care conversation. On a broad scale, standardization has benefits and arguably is the only proven way to improve care for large groups of people. Yet, as these examples show, standardization may harm highly able patients and shackle exceptional physicians.
One problem, of course, is that people tend to overestimate their abilities and assume they should receive the most complex, effective treatment strategies in every case. Perhaps the answer is to create explicit treatment levels—perhaps graded by high, medium, and low intensity or difficulty—and then allow patients and doctors to choose explicitly among them. That way, at least we’re acknowledging how health care works in the real world.
