Reducing Inappropriate Prescribing: Evidence from a Randomized Intervention in Medicare

There exist wide variations in practice patterns across health care providers. At the extremes these variations may reflect low or even negative value care. In this randomized evaluation, high-volume prescribers of the antipsychotic quetiapine (branded Seroquel) were sent peer comparison letters stating that they were under review. We find that the letters reduced prescribing of quetiapine by 11.1% over 9 months and 15.6% over 2 years. Receipt of quetiapine decreased more for patients with low-value indications than for patients who met quetiapine prescribing guidelines. There was little evidence of gaming by prescribers and we detected no patient harms in health care utilization data. The results show that peer comparison letters can raise the quality of prescribing, but more research is needed on how to effectively target low-value and unsafe prescription drug utilization.

There exist wide variations in practice patterns across health care providers. At the extremes these variations may reflect low or even negative value care. In this randomized evaluation, high-volume prescribers of the antipsychotic quetiapine (branded Seroquel) were sent peer comparison letters stating that they were under review. We find that the letters reduced prescribing of quetiapine by 11.1% over 9 months and 15.6% over 2 years. Receipt of quetiapine decreased more for patients with low-value indications than for patients who met quetiapine prescribing guidelines. There was little evidence of gaming by prescribers and we detected no patient harms in health care utilization data. The results show that peer comparison letters can raise the quality of prescribing, but more research is needed on how to effectively target low-value and unsafe prescription drug utilization.