Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Nudging Providers to Curtail Dangerous Opioid Prescribing: A Trial to Investigate Mechanisms
Last registered on October 21, 2020


Trial Information
General Information
Nudging Providers to Curtail Dangerous Opioid Prescribing: A Trial to Investigate Mechanisms
Initial registration date
October 20, 2020
Last updated
October 21, 2020 8:36 AM EDT

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Columbia University
Other Primary Investigator(s)
PI Affiliation
PI Affiliation
University of Southern California
Additional Trial Information
In development
Start date
End date
Secondary IDs
Despite an enormous policy response, opioid prescribing remains well above historical levels and harms from opioids continue to mount. Nearly all states have Prescription Monitoring Programs (PMPs) to facilitate safer prescribing of opioids and other drugs, but research suggests these systems only deliver benefits when health care professionals are required to use them. Even with PMP mandates in place, providers may be unaware of the dangers of co-prescribing opioids with benzodiazepines or gabapentinoids, which include increased risk of overdose and death. Working with the Minnesota state government, we will mail letters to guideline-discordant opioid prescribers that either highlight an upcoming legally mandated requirement to check the PMP before prescribing an opioid, inform and educate providers about their patients filling concurrent prescriptions and the dangers of such co-prescribing, or both. Study participants will be randomized to receive no intervention or one of the three treatment letters. Using administrative data, we will track effects of the letters on not only prescribing but also PMP usage and queries. Findings form the multiplicity of treatment messages and outcomes will shed light on the mechanisms driving overprescribing. Results will inform future work by state and local policymakers to make opioid prescribing safer.
External Link(s)
Registration Citation
Jacobson, Mireille, David Powell and Adam Sacarny. 2020. "Nudging Providers to Curtail Dangerous Opioid Prescribing: A Trial to Investigate Mechanisms." AEA RCT Registry. October 21. https://doi.org/10.1257/rct.6602-1.0.
Experimental Details
Prescribers will be assigned at random to one of four arms:
1. PMP Use Mandate Letter: Prescribers are sent letters with reminders about the mandate to check the PMP when prescribing opioids.
2. Prescribing Information Letter: Prescribers are sent letters with information about their patients who received co-prescriptions. The letters will provide clinical background on the harms of co-prescribing and encourage prescribers to avoid co-prescribing in the future.
3. Prescribing Information + PMP Use Mandate Letter: Prescribers are sent letters combining the content in arms 1 and 2, reminding them about the PMP use mandate and providing information about their patients who received co-prescriptions.
4. Control (As Usual / No Letter): These prescribers are not sent letters, the as-usual case.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
1. Checking the PMP during the 60 days after letters were sent
2. Avoid writing any co-prescriptions during the 60 days after letters were sent
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
At the start of the study, we will conduct a review of PMP records to identify prescribers who wrote opioid-benzodiazepine and opioid-gabapentinoid co-prescriptions. Prescribers will be randomized at 1:1:1:1 ratio to one of three treatment arms or a control arm. Then prescribers in the treatment arms will be sent two letters, one month apart.
Experimental Design Details
Not available
Randomization Method
Stratified randomization done using random numbers generated in Stata
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
12,000 prescribers
Sample size: planned number of observations
12,000 prescribers
Sample size (or number of clusters) by treatment arms
3,000 prescribers per arm
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Binary outcomes, all treatment arms vs. control: 2.9 percentage points (assuming no power gain from stratification and statistical controls), 1.5 percentage points (assuming 75% reduction in variance from stratification and statistical controls) Binary outcomes, arm vs. arm: 3.6 percentage points (no power gain), 1.8 percentage points (75% reduction in variance)
IRB Name
National Bureau of Economic Research
IRB Approval Date
IRB Approval Number
IRB Name
Columbia University Human Research Protection Office
IRB Approval Date
IRB Approval Number