Abstract
Background and Objectives: Opioid misuse has become a national crisis. In response to the need for improved standards of opioid prescribing within medical practices, a university-based academic family medicine practice developed, implemented, and evaluated a series of initiatives to ensure appropriate opioid prescribing and support clinicians in the practice in providing optimal care.
Methods: In 2015–2018, the University of North Carolina Family Medicine Center developed and implemented a practice-wide opioid prescribing policy, created a controlled medication advisory board (CMAB), provided regular feedback to clinicians on opioid prescribing, and trained selected providers in management of opioid use disorder. The impact of these opioid stewardship initiatives on prescribing patterns, utilization of the CMAB, and provision of medications for opioid use disorder was evaluated using electronic health record data from 2015 to 2018 and chart audits.
Results: Between 2014 and 2018 the opioid prescribing rate per 100 patient visits decreased by 31% and the rate of concomitant use of benzodiazepines and opioids decreased by 56%. The CMAB received 117 referrals between 2015 and 2018, 60% of which resulted in recommended revision in the treatment plan.
Conclusions: Safe opioid prescribing is essential to mitigate the opioid crisis. An evidence-based standardized protocol, coupled with support for providers and patients, can reduce prescribing and improve patient safety, thereby enhancing the comprehensiveness and quality of patient care.
- Addictive Behavior
- Benzodiazepines
- Chronic Disease
- Chronic Pain
- Family Medicine
- Opioid-Related Disorders
- Opioids
- Patient Safety
- Primary Health Care
Introduction
In 2017, the opioid crisis in the United States was declared a national public health emergency as fatal drug overdoses reached 70,237—nearly double the rate for car crashes or gun violence.1,2 Overdose deaths were 9.6% higher in 2017 than 2016,3 and opioids continue to be the leading cause of death under the age of 50 years.4 Despite acknowledgment of this crisis and overall decreased prescribing, opioid use has not abated.
While prescribing patterns are only 1 contributing factor, national organizations have urged providers to better regulate opioid prescribing5⇓–7. Recent years have seen a decline in prescribing from 81.3 prescriptions in 2012 to 58.5 prescriptions per 100 persons in 2017.6⇓⇓⇓–10 However, decreasing prescriptions alone is not sufficient to combat the crisis. Closely monitored tapering of opioid dosing and risk mitigation strategies are essential for patient safety and compliance. This need is demonstrated by evidence that rapid discontinuation of opioids increases the risk of opioid-related emergency department visits and hospitalizations.11,12
With the goal of following best practices for opioid prescribing, an academic family medicine practice in North Carolina developed a comprehensive opioid stewardship initiative, including practice-wide policies, ongoing prescription monitoring, and a controlled medication advisory board (CMAB). This brief report describes that initiative, its impact on practice patterns, and resources that may assist other primary care practices in addressing the opioid epidemic.
Methods
Practice Setting
The University of North Carolina (UNC) Family Medicine Center (FMC) is a clinical practice of 40 faculty and 28 resident physicians in central North Carolina. The clinic cares for a population of over 20,000 patients, with over 65,000 outpatient visits per year. Approximately 61% of patients have private insurance, 14% Medicare, 13% Medicaid, and 12% are uninsured.
Opioid Policy Development
In 2015, the UNC FMC formed a committee tasked with the following objectives: 1) develop a policy for prescribing chronic opioid therapy (COT), 2) update the existing treatment agreement, and 3) support providers and patients in the adoption of these new policies.
The committee included resident physicians, medical assistants, nurses, front-desk associates, a social worker, a psychologist, and faculty physicians. Although invited, the clinical pharmacist was unable to participate due to a job transition. The committee reviewed the existing opioid policy, policies of surrounding primary care clinics, literature, and multiple guidelines.7,13⇓⇓⇓⇓⇓⇓⇓⇓–22 Input was solicited from faculty, residents, and staff, and the final policy was approved by clinic leadership.
The policy (Appendix A) included recommendations to limit opioids to no more than 120 morphine milligram equivalents per day (MME/day), prescribe naloxone for any patient taking 50 mg or more MME/day, and avoid concomitant benzodiazepine use. The policy included appendices regarding how to calculate the MME/day, order and interpret urine toxicology results, successfully taper opioids and/or benzodiazepines to safer doses, navigate conversations with patients, and nonpharmacologic pain management resources.
The committee created scripts (Appendix B) to describe the new policy to patients and answer questions. To enhance documentation consistency several templated phrases were created in the electronic health record (EHR) (Appendix C).
Controlled Medication Advisory Board
In 2015, a board was created to support providers and patients with challenges adhering to the policy. The board was composed of 2 physician faculty, a social worker, a nurse, a psychologist, rotating resident physicians, and pharmacist when available. Providers and staff were encouraged to refer patient cases to the board (Appendix D). The board met monthly and referring providers were invited to attend.
The controlled medication advisory board (CMAB) reviewed each referral and conducted a chart review, including diagnoses, prior evaluations and treatment, medications, assessment of patient’s functioning, treatment agreement, and frequency and outcome of risk mitigation strategies, such as UDS and queries of the state’s Prescription Drug Monitoring Program. Risk considerations such as the patient’s psychiatric and substance use history, Opioid Risk Tool score, medication interactions, and current MME/day were also included.23 Finally, the board reviewed safety considerations such as prescribing naloxone or EKG monitoring for patients prescribed methadone. The process often uncovered patients who were at high risk for accidental overdose and some with histories consistent with opioid use disorder (OUD). The criteria for OUD are outlined in Table 1. The patient was not seen by the board.
Evaluation Methods
UNC FMC incorporated opioid management as a quality improvement goal. Evaluation included analyzing clinical and quality improvement data for 2015 to 2018 from UNC’s EHR. Baseline data were from January 2015 before implementation of the policy.
To identify the impact of the opioid policy on patients’ decisions to continue to receive care in the practice, the entire cohort of patients on chronic opioids (n = 856) in January 2015 was randomly ordered and the first 100 patients were selected for chart review. This cohort of 100 patients was followed through 2018 using the EHR and NC Prescription Drug Monitoring Program.
To describe the activities and outcomes of the CMAB, all referrals between 2015 and 2018 were reviewed and summarized.
The study was considered exempt by the UNC institutional review board as a quality improvement study with retrospective chart review.
Results
Changes in Opioid Prescribing
Table 2 shows changes in prescribing patterns at the UNC FMC between 2015 and 2018. The opioid prescribing rate per 100 patient visits decreased from a peak of 12.3 to 9.3 and benzodiazepine prescribing declined from 4.2 to 3.0. The number of patients on COT decreased by 31% and the number of patients on both chronic opioids and benzodiazepines decreased by 56%.
Of the cohort of 100 patients on COT retrospectively enrolled in January 2015, 74% remained active patients in 2018 and 2% followed a resident to a nearby clinic on graduation. Eight percent died of known causes unrelated to opioids, and no patients died of unknown causes or opioid overdose. Six percent left the clinic after their opioids were decreased or discontinued. None of those patients were discharged from the clinic. Ten percent were lost to followup; none of these patients had a decrease or discontinuation of their opioid regimen before leaving the practice.
Outcomes of the CMAB
Between 2015 and 2018, 117 referrals were submitted to the board. Requests generally fell into 2 domains: 1) patients not in compliance with the new policy and for whom an exception to the policy may be warranted, and 2) patients whose providers wanted help in changing their treatment plan to reduce risk or comply with the policy. Other requests included assistance tapering opioids and questions about whether opioids should be started in high-risk patients. Fifty-five providers made referrals.
Table 3 details the Opioid Risk Tool scores for referred patients and the outcomes from the consultations. Sixty percent of consultations resulted in recommendations for treatment plan revision. Common recommendations were to taper or discontinue opioids, taper or discontinue benzodiazepines, conduct further evaluation of pain, more frequent UDS, prescribe naloxone, or refer the patient to adjunctive therapies including behavioral health. Three percent of consults resulted in exemptions to the policy due to well-documented improvement in overall patient functioning and limited risk factors. A decrease or discontinuation of opioids was recommended in 39 patients, and follow-up chart reviews indicated 64% (n = 25) of those patients successfully decreased or discontinued opioids. Of the 33 patients for whom naloxone was recommended, 79% (n = 26/33) were subsequently prescribed naloxone.
Discussion
Developing and enacting a comprehensive approach to opioid prescribing led to several lessons. Involving each member of the care team in policy development facilitated broad policy acceptance. Providing support for providers and patients throughout the policy change process was crucial. Helping providers navigate challenging conversations, learn about alternative treatments, decide how and when to taper, and recognize potential signs of OUD was essential. The advisory board provided support, which is particularly important for large practices that include learners.
Prescribing opioids safely includes prescribing naloxone for patients at risk and avoiding prescribing opioids with benzodiazepines.5,6,22 While the data show that the number of patients taking either chronic opioids or concurrently prescribed opioids and benzodiazepines decreased, the number did not decrease to zero. Further research is needed to determine patient-centered strategies to discontinue 1 or both medications.
Increased monitoring helps providers identify patients who are misusing opioid medication. When misuse is identified, many clinics apply a “1 strike, you are out” rule, discontinuing all controlled medications or discharging the patient. However, misuse may indicate a diagnosis of OUD, a treatable medical condition. Instead of discharging patients, clinicians should offer treatment or referral for treatment of OUD.21
While it is easier to avoid starting a patient on chronic opioids, it is more difficult to help patients safely and successfully taper opioids as well as take opioid medication appropriately.18⇓–20 A comprehensive opioid stewardship initiative can provide prudent treatment for patients with chronic pain while increasing the resources and support available to providers.
Acknowledgments
The authors would like to acknowledge their colleagues at Piedmont Health Services for their contributions to the University of North Carolina Family Medicine Center controlled medication policy.
Appendices
I. Appendix A: UNC FMC Policy for treating Chronic Non-Cancer Pain with Opioids
II. Appendix B: UNC FMC Scripts provided to medical, nursing, and front desk staff to help them answer questions about the controlled substance policy
III. Appendix C: UNC FMC Dot Phrases or Quick Text for Documentation in the Electronic Medical Record
IV. Appendix D: UNC FMC Controlled Medication Advisory Board Referral Form
Appendix A: UNC FMC Policy for treating Chronic Non-Cancer Pain with Opioids
UNC Family Medicine Center Long Term Opioid Policy
Treatment of Chronic Non Cancer Pain with Long Term Opioid Therapy July 2018
This policy applies to chronic opioid use only. For short term opioid prescriptions for acute pain, no provider will provide more than 5 days of opioid pain medication which is in compliance with North Carolina law.
Background
Safe and effective treatment of chronic non cancer pain (CNCP) with long term opioids requires a team-based approach. When all members of our team follow these guidelines we can promote the cautious and selective prescribing of opioids while continuing to provide this treatment option for patients when the benefits outweigh the risks. In doing so we will promote the safety of our patients.
For some existing patients, the most appropriate treatment plan may not conform to these guidelines. Those cases will be referred to the Controlled Medication Advisory Board (CMAB, see below) for review for an exception to the policy.
Policy
1. Providers will prescribe opioids cautiously – using safer alternatives first and documenting what has been tried and failed. When opioids are indicated, providers will begin with low-dose, short-acting preparations, and make the decision to extend treatment beyond a trial period only after careful evaluation and discussion of benefits, harms, adverse events.
2. Providers will reduce risk of harm by carefully selecting patients who are candidates for long-term opioid therapy by doing a thorough risk assessment.
a. All patients will be screened for depression
i. Providers will work together with patients to pursue mental health treatment if indicated
ii. Patients with significantly uncontrolled mental health conditions are likely not good candidates for long term opioid treatment. They are considered higher risk and warrant very careful monitoring if opioids are prescribed.
b. All patients will be screened for alcohol and drug abuse
i. Patients who screen positive or for whom there is concern about potential substance misuse should be referred for a more comprehensive substance use assessment. UNC’s Alcohol Substance Abuse Program (ASAP) is one option for this type of evaluation. In addition, care managers can help refer patients for substance abuse evaluation within the Family Medicine Center.
ii. Patients with active alcohol and substance abuse are excluded from long term opioid therapy (including marijuana use).
iii. Patients with h/o significant alcohol or substance abuse, in remission, are likely not good candidates for long term opioid treatment. They are considered higher risk and warrant very careful monitoring if opioids are prescribed.
iv. Patients with a strong family history of alcohol and substance abuse may not good candidates for long term opioid treatment.
c. The NC controlled substances reporting system will be consulted, urine drug screen performed, and opioid risk tool administered annually (at a minimum) for all candidates for long-term opioid therapy.
d. The above risk assessment can be completed by providers, staff members as well as care managers
3. Providers will reduce risk of harm by restricting prescribing of high risk drugs, doses, and drug combinations.
a. maximum total opioid dose of 120 morphine milligram equivalents (MME) may be prescribed
i. Current patients receiving at or over 120 MME will discuss the new policy with their PCP. Providers will 1) start a slow taper of medication to reach the goal of 120 MME or 2) make a referral to a pain specialist or 3) request a policy exemption from the CMAB.
b. Methadone will be a medication of last resort
4. Providers will not prescribe opioids to patients who are known to be taking benzodiazepines (regardless of prescriber of benzodiazepines)
a. For existing patients who have been treated chronically with both opioids and benzodiazepines, one or both medications will be tapered in order to reduce overdose risk, with the goal of discontinuing one of the agents within 12 months.
b. Use of certain sleep aids (e.g. Ambien, Sonata, Lunesta) and other sedatives also increase overdose risk and will be avoided. Marinol increases overdose risk. Trazodone and diphenhydramine increase the risk theoretically.
c. Providers will not prescribe opioids to patients who are actively using marijuana
i. If a patient has a positive urine screen for marijuana, it is recommended that the need to discontinue marijuana be discussed with the patient. A urine screen should be repeated approximately 60 days later (allowing at least 30 days for THC to leave the patient’s system). A second positive urine screen for marijuana would be considered a violation of the treatment agreement. Substance use treatment resources should be offered to patients who think they may need assistance discontinuing marijuana on their own.
d. Providers will prescribe Naloxone kits to patients at high risk for overdose
i. Patients on 50 MED or higher
5. To ensure patient safety, long-term opioid therapy will include the following:
a. All patients receiving treatment for chronic non-cancer pain will have an identified prescribing PCP.
b. All patients requiring treatment for chronic non-cancer pain will have a meaningful assessment including complete history, physical, work-up of etiology of pain. The diagnosis resulting in chronic pain will be documented on the problem list, in addition to the diagnosis ‘Controlled Medication Agreement’.
c. All patients will have a treatment plan and treatment goals documented in the chart.
d. All patients will review and sign a treatment agreement for opioid therapy.
e. The treatment agreement will be scanned into the electronic medical record.
f. The agreement should be renewed annually or in the event of a change of primary care provider.
g. All patients will have regular evaluation of progress toward goals and side effects of treatment.
h. All patients will have periodic monitoring for adverse effects and signs of abuse, misuse and diversion. Frequency of monitoring will depend on risk assessment.
i. All patients will agree to information sharing with other medical professionals
ii. All patients will have a single listed prescriber for controlled medications. A single pharmacy is recommended
iii. All patients will bring their medicines (including pain medication) to the clinic for pill counts at every visit, and will bring all pain medications in to clinic for random pill count within 72 hours of being asked
1. Medications must be counted in front of the patient. The patient should be able to see the medications at all times. Never remove controlled medications from the exam room, and do not dispose of them for the patient
iv. All patients will agree to random urine drug screens (UDS)
6. Residents may not prescribe long term controlled substances to Department of Family Medicine employees or Family Medicine Center employees
7. Controlled medications should not be prescribed on the patient’s first visit to the FMC. Providers are encouraged to request previous records and review them prior to prescribing opioids for new patients. When possible, new patients requesting controlled medications should be scheduled for a double slot
8. When a provider leaves the department or a resident graduates, any patients treated by that provider will be transferred to another provider for evaluation
9. The provider who is assuming care of the patient will fully evaluate the patient and may elect to continue with or modify the treatment plan.
10. Providers on extended leave will partner with a covering provider to care for their patients with Controlled Medication Treatment Agreements. Covering providers may elect to change a patient’s regimen if necessary.
1. The Family Medicine Center will establish and maintain a Controlled Medication Oversight Committee (CMAB)
a. To advise the clinical team regarding:
i. Behavior problems that impede relationships or efficiency of the practice
ii. Substance abuse, diversion, medication misuse
iii. Pharmacologic questions in management of pain
b. To serve as an oversight body for the entire pain program to assure:
i. Consistent application of standards
ii. Safe prescribing practices
iii. Timely updates and improvements in the program as required
iv. Evaluation of cases that may be eligible for exception to the protocol
11. Patients with a Controlled Medication Treatment Agreement may require treatment by other providers for acute pain. This should be limited to urgent situations (unexpected surgery, fractures). Providers who prescribe a controlled medication in this situation will send a message to the patient’s PCP.
12. Providers are encouraged to use Epic dot phrases to aid in documentation
a. HPI for chronic pain: hpichronicpain
b. Assessment and Plan for chronic pain: apchronicpain
Appendix B: UNC FMC Scripts provided to medical, nursing, and front desk staff to help them answer patient concerns about controlled medications New Patients
1. For new patients, when asked about the nature of the visit, the response “pain” or unspecified “prescription refills” should trigger the following.
2. Explain: “As a new patient we want to provide the information you will need for your visit related to the prescribing of controlled substances. Initially, the provider will be getting to know you and collecting information about you and no prescriptions will be given for controlled medications at the first visit.”
When asked about the new policy, signing a treatment agreement, completing a urine toxicology screen, etc.:
1. Explain: “In our effort to provide safe and high-quality patient care, the UNC Family Medicine Center has developed new guidelines around the use of prescription pain medicine for the treatment of acute and chronic pain. We have also revised our guidelines for the prescribing of other controlled medications such as benzodiazepines for anxiety and stimulants for attention deficit disorder.”
2. “All of our new guidelines are based on increasing knowledge about the risks and benefits of these medicines, and on recent guidelines issues by the North Carolina Medical Board.”
When prescriptions run out as scheduled and the provider is unavailable for the return visit:
1. Explain: “Your provider is unavailable until *** but we can schedule you to be seen with another provider who will attempt to contact your provider about your prescriptions.”
2. When prescriptions run out early and the provider is not in clinic:
Explain: “Your provider is not available today, however I can schedule you to be seen by a colleague who will attempt to contact your provider about your prescriptions.”
1. When the patient asks, “Will I get my prescriptions?”:
2. Explain: “You will discuss this with the provider at your visit. They will make every attempt to consult with your primary provider, but I cannot guarantee that you will receive your refills.”
Appendix C: UNC FMC Dot Phrases or Quick Text for Documentation in the Electronic Medical Record
Note for Patients with Chronic Pain on long term opioids:
Subjective:
• Chronic Pain is caused by: ***
• Medications that patient is currently taking: ***
∘ Side effects?
• Nausea
• Vomiting
• Pruritis
• Constipation
• Dizziness/lightheadedness
• drowsiness
• confusion
• ***
• Physical function: Is this medication improving pt's function? {YES/NO:22418}
• How?
• Other modalities patient is using to control pain:
∘ hot/cold compresses
∘ hydrotherapy (shower, bathing)
∘ massage
∘ physical therapy
∘ chiropractic treatment
∘ TENS unit
∘ meditation
∘ pain psychologist
∘ group therapy
∘ guided imagery
∘ breathing exercises
∘ Acupuncture
∘ cognitive behavioral therapy
∘ yoga or other gentle exercise/stretches
∘ home exercise program
∘ injections or surgical tx
• referral to PM&R
∘ referral to pain clinic
∘ ***
• Any medication left over today? {YES/NO:22418}
• Any illicit substance use since last visit? {kbillicit:23315}
• Any concerning medication-related behaviors? {YES/NO:22418}
• When was the last dose of pain medication taken? Will it be in urine today? ***
Pain, Enjoyment and General Activity:
What number describes pain on average in the last week? {kb1-10:30530}
What number describes how pain has interfered with your enjoyment of life in the last week? {kb1-10:30530}
What number describes how pain has interfered with your general activity in the past week? {kb1-10:30530}
Assessment and Plan
• Pain is/is not adequately controlled with current plan
• Functional improvements that justify chronic opioid medications: ***
• Patient's plan currently includes: {kbpaintxoptions:30280}
• In the past patient has tried: {kbpaintxoptions:30280}
• Any aberrant behaviors? {YES/NO:22418}
• If on high dose opioids, does this pt have naloxone at home and know how to use it? {YES/NO:22418}
• Treatment agreement was signed by pt and myself on: ***
• PHQ -9 done: *** Result: ***
• Risk assessment done? Date***
Opioid Risk Tool Score: ***
• Refills given for *** months, dates: ***
• PDMP Database appropriate and last checked on: ***
Last Utox:
Appendix D: UNC FMC Controlled Medication Advisory Board Referral Form
Consult Request - Family Medicine Controlled Medication Advisory Board (please complete as much information as you can)
1. Reason for Consult:{Controlled Med Consult Reasons - FMC:37402:o}
2. Etiology of chronic pain:
3. Past diagnostic workup:
4. Past consultations:{Controlled Medication Past Consults - FMC:37407:o}
5. Complicating co-morbidities (include psychiatric diagnoses and history of substance misuse):
6. Non-opioid therapies tried:
7. Medication history and current meds/doses x duration of time on this regimen:
8. Functional assessment:
a. Pain, Enjoyment and General Activity:
i. What number describes pain on average in the last week?.................. {kb1-10:30530}
ii. What number describes how pain has interfered with your enjoyment of life in the last week?.................... {kb1-10:30530}
iii. What number describes how pain has interfered with yourgeneral activity in the past week?................. {kb1-10:30530}
9. Most recent NCCSRS repsort and recent urine tox screen results:
10. Any concerning aberrant behaviors?
11. Opioid Risk Tool - total score:
Risk Category:
Low Risk: 0 to 3
Moderate Risk: 4 to 7
High Risk: 8 and above
Notes
This article was externally peer reviewed.
Conflict of Interest: None.
Funding: None.
To see this article online, please go to: http://jabfm.org/content/33/6/992.full.
- Received for publication December 20, 2019.
- Revision received May 29, 2020.
- Accepted for publication June 2, 2020.