Back to History Current Version
The Effect of Informative Letters on the Prescription and Receipt of Schedule II Controlled Substances
Last registered on January 20, 2015


Trial Information
General Information
The Effect of Informative Letters on the Prescription and Receipt of Schedule II Controlled Substances
Initial registration date
October 06, 2014
Last updated
January 20, 2015 8:59 PM EST
Primary Investigator
Other Primary Investigator(s)
PI Affiliation
General Services Administration
PI Affiliation
Harvard University
Additional Trial Information
On going
Start date
End date
Secondary IDs
Fraud and waste is estimated to cost the American health care system nearly $200 billion each year, and the public Medicare and Medicaid programs about $60 billion each year. This study will evaluate a new method for fighting fraud: mailing informative letters to outlier providers to notify them of their aberrant behavior. These letters are targeted at high prescribers of schedule II controlled substances in Medicare Part D. We will look at the effects of these letters on the behavior of providers and their patients. These effects are of substantial policy interest as they suggest how to best design anti-fraud policies. They are also of academic interest, shedding light on the behavior of physicians and their patients.
External Link(s)
Registration Citation
Finkelstein, Amy, Adam Sacarny and David Yokum. 2015. "The Effect of Informative Letters on the Prescription and Receipt of Schedule II Controlled Substances." AEA RCT Registry. January 20.
Experimental Details
A letter to outlier prescribers of schedule II controlled substances explaining that their prescribing behavior is highly unlike their peers.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
An analysis was conducted to identify outlier prescribers of Schedule II controlled substances in the Medicare Part D. This analysis identified a group of prescribers, and they were then randomly allocated to a treatment or a control group.
Experimental Design Details
Not available
Randomization Method
Randomization done by computer (in Stata)
Randomization Unit
The prescriber (e.g. physician, nurse, physician assistant, etc.)
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Approximately 1500 prescribers
Sample size: planned number of observations
Approximately 1500 prescribers
Sample size (or number of clusters) by treatment arms
Approximately 750 providers in the control arm and 750 providers in the treatment arm
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We constructed a sample of outlier prescribers that replicated the method used in the study but was built from a more limited dataset. Using these prescribers we estimated that at a significance level of 5% and a power of 80%, we could detect a change in the number of schedule II prescription drug events of 136.9 (baseline mean level: 1,556.7; standard deviation: 1,231.2; effect as a share of baseline mean: 8.8%) and a change in the dollar value of schedule II prescriptions of $34,069 (baseline mean level: $217,204; standard deviation: $239,759; effect as a share of baseline mean: 15.7%). Since we will have access to richer data with better control variables, we believe that these estimates are conservative.
IRB Name
MIT Committee on the Use of Human as Experimental Subjects
IRB Approval Date
IRB Approval Number
IRB Name
Harvard Committee on the Use of Human Subjects in Research
IRB Approval Date
IRB Approval Number