Behavioral nudges lead to striking drop in prescriptions of potent antipsychotic

"Compared with existing work on prescribing quality, our study provides a unique example of a large-scale, low-cost intervention yielding clinically meaningful and long-lasting effects," said Adam Sacarny, PhD, assistant professor of Health Policy and Management at Columbia's Mailman School of Public Health and first author. The work was conducted with the Office of Evaluation Sciences in the General Services Administration, which promotes evaluation and testing in the federal government; the Centers for Medicare and Medicaid Services, which administers the Medicare program; and researchers at the Harvard T.H. Chan School of Public Health.

The study was a randomized controlled trial targeting the 5,055 highest Seroquel-prescribing primary care physicians nationwide in the Medicare Part D (prescription drug coverage) program in 2013 and 2014. A random half of the doctors were assigned to the treatment arm and received 3 letters comparing their prescribing practices to their peers; the other half received placebo letters about an unrelated Medicare regulation. The treatment arm's letter stated that the physician's prescribing of quetiapine high relative to their peers was under review. The text also discussed that high quetiapine prescribing could be appropriate but was concerning for medically unjustified use. The letter encouraged primary care physicians to review their prescribing patterns.

The physicians who got the peer comparison letters dropped their overall Seroquel prescribing by 11 percent over the next 9 months and 16 percent over the next 2 years. New initiations of Seroquel dropped even more: 24 percent over 2 years.

"Our findings show that for health care organizations and clinicians aiming to improve the quality of prescribing, peer comparison messages could be useful and effective tools, particularly when they are paired with a review of previous prescribing activity," said Sacarny. "Similar messages could address over-prescribing of other drugs, like opioids, or they could target care that goes against clinical guidelines."

The researchers also followed patients who had been treated by the physicians in the study. On average, patients of the treatment group physicians received 6 percent fewer days of Seroquel from all prescribers over 2 years. Patients with a history of dementia, where guidelines discourage Seroquel prescribing, experienced a larger reduction of 8 percent fewer days of Seroquel. There was no evidence of adverse effects on patients due to the letters: use of the emergency department, hospitalizations, and mortality were similar for patients of treatment and control physicians.

Seroquel is frequently prescribed "off-label," or outside its FDA-approved indications, for patients with dementia or Alzheimer's disease despite a large body of evidence that it is associated with significant harm in these populations. Guidelines from the American Psychiatric Association and the American Geriatrics Society discourage this overuse of antipsychotics including Seroquel. This off-label use against guidelines has attracted the attention of the Centers for Medicare & Medicaid Services (CMS) and federal oversight agencies.

"Overprescribing of antipsychotic drugs is a huge and persistent problem, particularly in the elderly. We think our results are striking for something as simple and cheap as sending letters," said Dr. Michael Barnett of the Harvard T.H. Chan School of Public Health, the second author on the study. "Beyond the benefit in reducing anti-psychotic use, the intervention could also have implications for how we can nudge physician behavior more broadly."

Sacarny and colleagues showed that with the increasing need to address the dangers of inappropriate prescribing, peer comparison letters targeted at high-risk medications provide one way to efficiently create lasting changes in prescribing patterns.