GSA’s Office of Evaluation Sciences (OES) teamed up with the Centers for Medicare and Medicaid Services (CMS) and researchers from Columbia University, Massachusetts Institute of Technology, and Harvard University, to try to discover an impactful, low-cost solution to reduce overprescribing of prescription drugs.
In two studies, OES conducted two randomized evaluations to measure the impact of peer comparison letters on prescription rates. The evaluations build off an existing CMS program that provides physicians with data on Medicare billing trends and allows recipients to compare their billing practices to their peers. Findings from the second study were recently published in the August 2018 issue of the journal JAMA Psychiatry, as well as on the OES website.
In the first study, previously published in Health Affairs in March 2016, CMS sent a peer comparison letter to randomly selected high-volume prescribers of Schedule II drugs (controlled substances such as opioids and amphetamines), indicating that their prescribing was extremely high relative to their in-state peers. Using administrative claims data from the 90 days after the letter was mailed, comparisons failed to detect an effect of the letter on subsequent Schedule II prescribing.
Given the low cost of letter interventions, and the fact that peer comparison letters have been shown to work in other contexts, CMS decided to conduct a second study.
In the second study, the prescription of interest was quetiapine (branded Seroquel), an antipsychotic which is often prescribed for reasons not supported by clinical evidence. Quetiapine was of particular interest because although the U.S.. Food and Drug Administration (FDA) warns against prescribing antipsychotic drugs for patients with dementia, the U.S. Government Accountability Office (GAO) estimates that about one-third of nursing home residents with dementia were prescribed an antipsychotic drug. Quetiapine was the most popular, accounting for 49 percent of the prescriptions and $158 million Medicare Part D payments in 2012. As a result, since 2012 CMS has initiated a number of strategies to reduce the use of unnecessary antipsychotic medications like quetiapine in nursing homes.
The second study built on the first by emphasizing oversight of potentially inappropriate prescribing. Letters were sent to prescribers multiple times, relied on more recent administrative data on prescribing rather than data from previous years, and used more accurate prescriber mailing addresses. CMS sent the series of peer comparison letters to randomly selected high-volume prescribers of quetiapine, stating that their prescribing was extremely high relative to their peers and that it was under review.
In this study, prescribers who were sent the peer comparison letters supplied 11.1 percent less quetiapine in the nine months following the intervention, compared to the control group. The patients of prescribers who were sent peer comparison letters also received less quetiapine than patients of the prescribers who received a control letter. While the reduction was 3.9 percent for the average patient, it was larger for the subset of patients who did not have FDA-approved indications for quetiapine: Those patients saw a 5.9 percent reduction in quetiapine receipt following the intervention. This result held up over time: two years after the intervention ended, the lower prescribing and receipt of quetiapine remained. Furthermore, there was no evidence of substitution to other antipsychotic drugs, and mortality and hospital utilization were similar in the treatment and control groups after nine months.
Mona Siddiqui, Chief Data Officer for the Department of Health and Human Services (HHS), said, “this collaboration between the Office of Evaluation Sciences and the Centers for Medicare and Medicaid Services has helped HHS build evidence about what works to address inappropriate prescribing of strong prescription drugs like antipsychotics. The use of administrative data for two rounds of rapid cycle testing meant HHS quickly found a low-cost intervention that produced meaningful results.”
Considering the cost to Medicare Part D of antipsychotics like quetiapine and concerns about over-prescribing of these and other drugs, low-cost interventions like this one offer the potential to reduce spending in Medicare while raising the quality of health care that beneficiaries receive. The repeated collaboration between OES and CMS and use of existing CMS administrative claims data allowed for iterative testing and learning across two studies. The result was a high-quality evidence base about building low-cost, light-touch interventions with the potential to improve health care safety and reduce health care costs.
Learn more about OES and explore their results.