Commentary: Small Changes Can Help Reduce Opioid Prescriptions

A major source of the opioid crisis is overprescribing by well-meaning doctors who want to relieve patients’ pain, but are insufficiently focused on the risks. Could behavioral economics help change that — and save lives?

Almost 2 million Americans are now addicted to opioids. In 2016 alone, 42,000 died from opioid overdoses, with more than 32,445 deaths coming from prescription opioids. These numbers reflect alarming increases in a short period. In 2015, for example, the number of deaths from prescription opioids was 22,598.

What can be done? Led by the University of Southern California’s Jason Doctor, a team of researchers found a dramatic way to nudge doctors to reduce opioid prescriptions. Their starting point was simple: When patients die, clinicians often don’t find out.

Their experiment involved 861 clinicians in San Diego. About half of them served as the control. The other half received this letter: “This is a courtesy communication to inform you that your patient (Name, Date of Birth) died on (date). Prescription drug overdose was either the primary cause or contributed to the death.” The letter also offered information on the risks of prescription medication-related deaths. It suggested consultation of a website with additional materials.

The researchers hypothesized that the letter would reduce opioid prescriptions. To test that hypothesis, they compared the number of opioid prescriptions a few months before and a few months after the letter was sent.

In the control group, prescriptions stayed pretty steady (actually they increased modestly). In the group of clinicians that received the letter, by contrast, prescriptions decreased significantly. And those clinicians were less likely to start new patients on opioids at all.

Why did that happen? The researchers speculate that like everyone else, clinicians assess health and safety risks by asking whether bad outcomes come readily to mind. They use the “availability heuristic,” as Daniel Kahneman and Amos Tversky call it. The letter made a tragic outcome quite salient. That affected clinicians’ behavior.

People are also likely to act differently when they are being watched, especially by those in positions of authority. A letter from the San Diego medical examiner about a patient’s death is bound to get a doctor’s attention.

The researchers emphasize that their approach could be used in many cities and states. In view of the magnitude of the opioid crisis, that’s an excellent idea.

True, we don’t want to deter doctors from prescribing opioids when they’re needed. But feedback about risks is important – and if it’s vivid, it is likely to have an impact.

The San Diego study follows on the heels of another successful effort to reduce opioid prescriptions, also informed by behavioral economics.

At two emergency departments in Philadelphia, the University of Pennsylvania’s Medicine Nudge Unit changed the default setting on electronic medical records discharge orders, from 30 or more tablets in the initial prescription of opioids to just 10.

You might not think that this change would have much of an effect. After all, a default setting is just a starting point — a mere nudge. Clinicians can do what they want.

But the change in the default had a real impact. It increased initial prescriptions of just 10 tablets by 22 percent (from 21 percent to 43 percent) — and significantly decreased larger initial prescriptions.

Here as well, there is a major opportunity. Electronic health records have many default settings, which influence clinicians’ decisions. Some of those settings haven’t been adequately thought through. A lower default for opioid prescriptions could reduce the level of addiction.

The two experiments turn out to be closely linked. Opioid prescriptions are sometimes a product of habitual behavior rather than careful consideration. Even though the opioid crisis is highly publicized, many doctors just keep doing what they have done before. Unless they receive feedback when tragedy strikes, or unless they are given a new default, they have no reason to change.

That’s a problem, but it’s also an opportunity. Especially when it is both concrete and personal, a little information can go a long way.

Many officials have been discouraged to learn that apparently promising efforts to limit opioid prescribing, including aggressive regulatory restrictions, do not appear to have accomplished much. The good news is that more modest approaches, informed by behavioral economics, are having a real impact.

There’s much more to learn. But we already know something important: In view of their life-saving potential, those approaches should be adopted more widely — and soon.