Of course, not every inappropriate medical practice has the same high stakes. Several researchers mentioned the overuse of M.R.I.s for lower back pain as the sort of thing that would be nice to change with a nudge. But no one dies of an M.R.I. overdose.
There are also risks that the strategy’s success could become its undoing. Kevin Volpp, the director of the center for health incentives and behavioral economics at the University of Pennsylvania, remembers how the first few pop-up alerts in electronic medical records helped guide doctors toward best practices. But as they proliferated, they became easier to ignore. He wondered if sternly worded letters might also have diminishing returns if they became more popular. “Once you have 30 reminders flying at you from every direction, they may not be so effective anymore,” he said.
It’s also unclear if the letters helped doctors avoid prescribing drugs to patients who could be harmed — or just reduced prescribing altogether. Both studies found bigger effects on prescriptions written for the doctors’ new patients. And the antipsychotic paper found that doctors wrote fewer prescriptions both for the target group — patients who would not benefit from the drugs — and those who, in line with their diagnoses, would have benefited. This suggests that a scary letter can be a blunt tool.
And, of course, even the perfect letter is not going to be enough to change medical practice on its own. “The world needs bigger changes than what you saw in the studies,” said Peter Ubel, a professor of business, public policy and medicine at Duke, who has studied behavioral approaches to medical care. “But what a nice start.”
Researchers said that nudges are exciting when they work because they are cheap. They are also easier to test than many other strategies. Both of the successful nudge letters were examined using randomized controlled experiments, the kind of study that is common in evaluating drug effectiveness, but rarely used to study health care delivery. The ease of testing tiny changes to a letter’s particulars means that researchers may be able to further refine their nudges over time.
“If you want to stop something from happening, you have to build up rocks by the river, along the edge, to dam it up,” said Jason Doctor, a co-author on the opioids prescribing paper, which was published in the journal Science last month. The doctors in his study cut back on a standard measure of overall opioid prescribing by 9.7 percent — not a huge decrease. But if such nudges were added to new medical guidelines and other tools, they could be part of a broader solution to the problem of overprescribing, he said.