Increasing Naloxone Co-prescription for Patients on Chronic
Opioids: a Student-Led Initiative
Jonathan E. Freise, BS, MS
, Elizabeth E. McCarthy, BS
, Scott Steiger, MD
and Leslie Sheu, MD
School of Medicine, University of California, San Francisco, San Francisco, CA, USA;
Department of Medicine , University of California, San
Francisco, San Francisco, CA, USA.
KEY WORDS: naloxone; opioids; primary care; quality improvement;
J Gen Intern Med 33(6):797–8
© Society of General Internal Medicine 2018
In 2016, the Center for Disease Control (CDC) established
guidelines for primary care providers prescribing opioids for
chronic pain, including naloxone co-prescription for patients
on high-dose chronic opioids (morphine equivalent daily dose
(MED) ≥ 50 mg).
Despite naloxone’s proven efficacy, naloxone co-
prescription in primary care settings remains low.
student-led quality improvement effort assessed barriers to
naloxone co-prescription at an academic primary care clinic
and implemented and evaluated a targeted intervention to
increase naloxone co-prescription for patients on high-dose
chronic opioids for non-cancer pain.
As part of a new medical school pre-clerkship curriculum,
first-year medical students were paired with faculty to conduct
quality improvement projects within the faculty members’
clinical setting. We aimed to increase naloxone co-
prescription for a registry of patients prescribed chronic opi-
oids within the University of California, San Francisco
(UCSF), Division of General Internal Medicine, an academic
primary care clinic.
In a needs assessment survey with multiple-choice ques-
tions, providers were asked about barriers to prescribe nalox-
one by selecting all answers that applied (out of eight possible
choices). Forty-nine of 117 (42%) providers responded, and
the two most commonly reported barriers were lack of pro-
vider comfort (41%) and lack of time (59%), consistent with
To increase provider comfort, we sent indi-
vidualized emails to all providers who had patients in the
registry (N = 101), including each patient’s MED, naloxone
co-prescription status, and a suggested script for discussing
naloxone with patients based on previous recommendations
To address time burden, we provided medical
assistants (MAs) with a list of patients on high-dose chronic
opioids without a naloxone co-prescription and an upcoming
clinic appointment every 2 weeks during the intervention
period (October 2016–February 2017). MAs queued naloxone
prescription orders in the electronic medical record (EMR)
during the check-in process to serve as visual reminders for
providers to discuss naloxone and to reduce the time for
prescribing naloxone. If desired, providers could complete
an order for Bnaloxone counseling^ to prompt licensed voca-
tional nurses (LVNs) trained in naloxone counseling to teach
patients how to use the medication. Patient education lasted 5–
10 min, including watching an informational video and a
demonstration using a sample of naloxone.
A sample individualized email sent to providers at the
How should I discuss a naloxone prescription with my patient?
A great time to bring up a naloxone prescription is while you are
checking in about pain management with your patient’s current pain
medication. Recent high-profile cases of accidental opioid overdoses
(e.g., Prince) can be a great starting point for talking about the dangers
of accidental overdose with opioids.
When suggesting a naloxone prescription, you can remind patients that
naloxone is a safety measure not just for them but also for family
members or friends who might take their opioids. It is helpful to use
phrases such as Bbad reaction^ and Bslows or stops breathing^ because
patients prescribed chronic opioids do not usually consider themselves
to be at risk of an overdose. This can also be a lead into educating your
patient about the dangers associated with chronic opioid use and
suggesting alternative pain management strategies.
One notable best practice for naloxone prescription is to suggest that an
at-risk patient create a Bbad reaction plan^ to share with friends,
partners, and/or caregivers. Such a plan would contain information on
the signs of overdose and how to administer naloxone or otherwise
provide emergency care (as by calling 911).
Our primary outcome was the proportion of patients with
naloxone co-prescriptions before and after our intervention.
We performed a negative binomial regression to determine the
effect of our intervention on the monthly number of naloxone
co-prescriptions in the clinic. We also report the number of
monthly naloxone co-prescriptions after completion of our
intervention. Institutional review board approval by the UCSF
Jonathan E. Freise and Elizabeth E. McCarthy contributed equally to this
Published online March 19, 2018