1

Opioid Prescribing Work Group

Minutes — December 14, 2015 12:30 – 3:30 p.m. 444 Lafayette Building, St. Paul

Members present: Julie Cunningham, Chris Eaton, Tiffany Elton, Rebecca Forrest, Ifeyinwa Nneka Igwe, Chris Johnson, Ernest Lampe (non-voting), Matthew Lewis, Pete Marshall, Murray McAllister, Richard Nadeau, Mary Beth Reinke (non-voting), Charles Reznikoff, Alvaro Sanchez, Jeff Schiff (non-voting), Matthew St. George, Lindsey Thomas

Members absent: None

State employees:

Department of Human Services: Charity Densinger, Sara Drake, Ellie Garrett, Tara Holt, Sarah Rinn

Department of Health: Dana Farley

Department of Labor & Industry: Lisa Wichterman

Board of Nursing: July Sabo

Guests: Shelly Elkington (Avenues for Care), Todd Gabrielson, Alexi-Reed Holtum (Steve Rummler Hope Foundation), Trudy Ujdur (Sanford Health)

I. Welcome and Introductions

Jeff Schiff called the meeting to order, and introductions were made around the room. Todd Gabrielson shared the compelling story of losing his daughter to an adverse reaction to codeine, which had been prescribed while she was hospitalized.

Ellie Garrett is working to schedule the January meeting and to set a standing meeting date and time. Schiff announced that Chris Johnson has offered to chair the group, and he expressed his appreciation to Johnson. Voting OPWG members will choose a chair at the next meeting, and members are encouraged to put forward their names if interested in serving.

A motion was made and seconded to create an ex-officio, non-voting seat for a representative of the Minnesota Department of Health. The motion passed unanimously.

II. Protocol Domains: Prescribing for acute pain

Garrett presented an overview of the goals for today’s meeting, which included voting on the protocol domains for prescribing opioids to treat acute pain and discussing the content of the first four domains: assessing and documenting (1) function, (2) pain, (3) physical health and risks and (4) mental health and substance abuse. She suggested that the work group use ICSI’s Acute Pain Assessment and Opioid Prescribing Protocol as a starting point for discussion and that members focus initially on primary care

2

prescribing. Modifications for other providers and sites of care (e.g., dental, surgical, emergency department) could be made later. A copy of Garrett’s brief presentation is available upon request from OPWG staff.

Garrett asked whether there were other prescribing protocols that the work group should review. Richard Nadeau agreed to circulate a copy of the Minnesota Dental Association’s protocol.

Schiff asked the committee to focus on the first four domains today. A member noted that nomenclature can create confusion. There is a difference between acute pain and treatment of pain in an acute care setting. Pain in an acute setting can be many things: mental health crisis, running out of medications, a flare-up of a chronic condition. This segued into a discussion about the importance and challenge of precisely diagnosing the patient. There are many instances in which pain may appear to be acute but may really be arising from a chronic problem. The committee agreed to focus initially on acute presentation of pain with no other confounding factors, such as chronic conditions.

Schiff asked the work group to consider the reorganized domains, as reflected in the version circulated for today’s meeting. The first eight, shaded items are considerations that come first: issues influencing whether to prescribe opioids; the remaining domains concern how to prescribe opioids, having decided that opioids are appropriate in the instant case. (See Attachment A.)

A member suggested moving item #9 (checking the Minnesota Prescription Monitoring Program) up to the shaded area. The PMP results are part of patient assessment. Others noted that checking the PMP can inform whether the patient really is presenting for a new, acute problem. He observed that the only objection might be that it is a hassle for the provider. Other members countered that the only hassle is finding something that looks concerning that requires more time. Physicians are allowed to delegate checking the PMP to another member of the treating team (though pharmacists on the team need better access to the PMP), and checking the PMP only takes a couple of minutes once doing so is routinized into the provider’s practice. Consensus emerged that discussions about checking the PMP should be moved up to the section on considerations prior to prescribing. A member stated that it is important to refer patients to counselling for addiction care if the patient assessment, including PMP findings, suggest that the patient might be suffering opioid use disorder. Another member agreed that there needs to be a follow-up in the protocol when there is a finding from the PMP.

A member stated that urine screening should also be considered, and identifies problems that do not always show up on the PMP. Another member countered that urine screens would be expensive and likely not needed for every patient. A member suggested that urine screening might be considered as a follow-up in the PMP.

A. Assessment of function

Garrett observed that there are many tools available for assessing function. Page 10 of Washington’s guidance contains some combined pain and function scales in the context of chronic pain that might be useful to think about earlier during acute presentation.

A member questioned whether assessing function is really relevant for treating acute pain. Another member agreed, noting that the main reason to address function is to establish therapeutic goals: The opioid’s purpose is to help the patient function better and not to eradicate pain completely. Well managed pain is part of the healing process.

3

A member suggested that assessing ability to perform activities of daily living (ADLs) is objectively verifiable information that can help assess healing post-surgery or trauma.

B. Assessment of pain

A member stressed that pain is not a vital sign. Vital signs require immediate, clinical response in order to address a symptom in the moment. Successful pain treatment should be measured in a month, not minutes or hours. There is enormous pressure on providers to lower their patients’ pain scores. Another member agreed and stated that patient satisfaction measures on pain actually lead to inappropriate and ineffective pain management.

Schiff suggested that concerns regarding measuring pain as part of patient satisfaction should be something that the committee considers in a future meeting. He cautioned, though, about exacerbating racial and ethnic disparities in pain treatment.

A member stressed the importance of setting realistic goals for improvements in relative pain scores. A member stated that pain scores are useful clinically. Another member observed that pain scores are subjective, and one what one patient terms a five might be another patient’s one or two or ten. Another member stated that the pain scale needs to be understood as reflecting not just pain but also emotional distress about pain. Schiff asked whether it would be reasonable to expect that the assessment and documentation include the patient’s observed nature, e.g., distractibility or stoicism.

A member suggested that pain scales are fine, so long as they are not used to measure provider performance or dictate clinical responses in the abstract without considering the patient’s full clinical presentation. Pain scales are only a guide, and should be used to help assess relative goals – like a 30 – 50% reduction in pain. Treat the patient, not the number. No numerical response to the pain scale alone warrants an opioid in and of itself.

A member suggested that the committee and/or DHS formally recommend to the Joint Commission that pain scores be removed from accreditation or other measurement standards. Several members agreed. Another member observed that corporate and system policies impede the appropriate prescribing of opioids. A member suggested that this topic is important enough for a separate agenda item, and he suggested tabling it for now.

Another member stated that cognitive threat and distress levels can be experimentally manipulated to help with diagnosis. Reporting the same sensation in the presence of different cognitive threats is instructive. Another member observed that the notion of “staying ahead of the pain,” while appropriate in the treatment of terminally ill patients, is harmful in other contexts.

A member observed that Washington’s guidance does a good job discussing the limits of opioids and their use. Opioids have limits, and there are more effective strategies for some patients and some conditions.

A member drew the group’s attention to ICSI’s guidance (page 9, section 3) on comprehensive pain assessment. Another member suggested that the risk/benefit grid page 19 of the ICSI guidance is also quite useful. Prescribers should document the physical findings that are consistent with pain and any prescription, including objective observation of discomfort, restlessness, tachycardia, crying, etc.

A member expressed concern about unintended consequences – that overemphasis on objective documentation could increase MRIs and other scans and tests inappropriately. Another member clarified

4

that objective assessment should document observations about the patient’s discomfort. Members discussed the utility of SOAP documentation criteria (subjective, objective, assessment and plan). Some prescribers are not documenting even the barest of justifications for prescribing.

Schiff brought the discussion to a close and asked the group to either endorse ICSI’s acute guidance vis-à-vis assessment and documentation or else propose concrete modifications. A motion was made and seconded that the OPWG recommend both objective documentation of the patient’s presentation of pain and diminished physical function. Documentation should include use of the pain scale as a relative tool, and concordance of the patient’s assessment of his or her own pain with the prescriber’s objective observations. The motion carried unanimously.

1. Public Comment

Trudy Udjur of Sanford Health (no conflicts of interest to disclose) stated that prescribing opioids well is complex. The pain scale is not diagnostic. A person can exhibit significant pain on the pain scale simply because they have run out of opioids.

C. Assessment of mental health risks

A member suggested that ICSI’s ABCDPQRS risk assessment approach (ICSI guidance, page 14) would be a useful place to start:

Alcohol use

Benzodiazepines and other drug use

Clearance and metabolism of the drug

Delirium, dementia and falls risk

Psychiatric comorbidities

Query the PMP

Respiratory insufficiency and sleep apnea

Safe driving, work, storage and disposal

Another member commented that safe driving will have to be addressed as part of the protocol at some point.

Schiff asked the group to focus on mental health and substance abuse risks for the moment. A member suggested that focus on anxiety disorders and depression would be a useful place to start. Another member stated that no screening tool is perfect, but revising past mental health history and medications is helpful. Even when a patient is presenting with a trauma, such as a femur fracture, knowing that the patient has a history of substance abuse or mental health disorders would be useful.

Members discussed the tension between adding to the prescriber’s burden with extensive documentation and history requirements for the initial prescription in cases of objectively verifiable pain vs. shifting some of the burden to the sub-acute prescribing timeframe. Taking a complete family history might be too burdensome as a standard for all initial prescriptions. A member stated that a bare minimum should include current or past addiction history, depression, anxiety, PTSD and suicidality.

A member suggested that risk is also addressed when prescribing a smaller dose for a shorter duration for the initial prescription. Prescribing too many pills is a very big problem.

5

Returning to the question of mental health and substance abuse, members discussed how comprehensive a history should be required in connection with an initial prescription. A member observed that current medication history is instructive, both to avoid concomitant use of dangerous combinations of drugs (such as benzodiazepines and opioids) and to inform the prescriber about current mental health conditions. Another member stated that asking about substance abuse history should be required at a minimum. It’s also important to ask patients if they feel like harming or killing themselves in order to assess suicidality.

Members discussed different sites of care, questioning whether the same standard should apply in primary care, dentists’ offices, the emergency department and so forth.

Schiff surveyed members about whether suicidality should be assessed in every setting for every initial prescription. Eight voting members said yes; five said no.

One member said that mental health history will capture suicidality. Another member clarified that any member of the treating team could take a history, so that should help with managing burden to the provider.

A member stated that if any questions about mental health or substance abuse are positive, then that should prompt further inquiry, including questions about past history of overdose.

Schiff polled members, and they agreed by a show of hands that the assessment should include a medication review and brief screening for substance abuse disorder and acute suicidality. Members briefly discussed that any positive findings could prompt a more thorough history, including questions about prior overdose or a decision to prescribe Narcan.

D. Assessment of physical health risks

Schiff drew members’ attention back to the ABCDPQRS assessment recommended by ICSI. Members discussed several risks, including advanced chronic obstructive pulmonary disease, renal failure, obesity and sleep apnea. Age cuts both ways, with youth being more associated with addiction risks and overdose more associated with middle and old age.

A member observed that using the ABCDPQRS assessment would support a culture change around prescribing. Another member added that opioids should never be prescribed, even to someone in obvious, traumatic pain, without informed consent about the risks, especially the risks of dependence and addiction. If prescribing to someone with a history of addiction because of severe injury or other objective, painful indication, then the prescriber should also refer the patient to an addiction specialist.

E. Next Steps

Members asked to discuss prescription dose and duration at the next meeting. They will also need to discuss prescribers and settings other than primary care, such as surgery, dentistry and emergency medicine.

In response to a question, Schiff stated that the Board of Pharmacy is likely to introduce legislation proposing mandatory enrollment in (as opposed to mandatory use of) the PMP.

Schiff adjourned the meeting at approximately 3:30 p.m.

6

Attachment A

1

Opioid Prescribing Work Group

Minutes — November 23, 2015 12:30 – 3:30 p.m. 444 Lafayette Building, St. Paul

Members present: Julie Cunningham, Chris Eaton, Tiffany Elton, Rebecca Forrest, Ifeyinwa Nneka Igwe, Chris Johnson, Ernest Lampe (non-voting), Matthew Lewis, Pete Marshall, Murray McAllister, Richard Nadeau, Mary Beth Reinke (non-voting), Charles Reznikoff, Jeff Schiff (non-voting), Matthew St. George, Alvaro Sanchez, Lindsey Thomas

Members absent: None

State employees:

Department of Human Services: Commissioner Lucinda Jessen, Cambray Crozier, CharityDensinger, Sara Drake, Ellie Garrett, Tara Holt, Melanie LaBrie, Sarah Linville, Sarah Rinn

Department of Health: Dana Farley, Mark Kinde

Department of Labor & Industry: Lisa Wichterman

Board of Nursing: July Sabo

Guests: Adam Fairbanks (Valhalla Place), Tom Freeman (Faegre Baker Daniels), Cara Geffert (HealthPartners), Charles Hilger (Valhalla Place), Juliana Milhofer (Minnesota Medical Association), Jeremy Olson (Star Tribune)

I. Welcome and context for the OPWG’s work

Jeff Schiff welcomed everyone to the first OPWG meeting, and brief introductions were made around the room. More detailed introductions of OPWG members were interspersed throughout the meeting. Commissioner Jesson addressed the meeting, welcoming the work group and guests. She stressed the importance of addressing the prescription opioid crisis and thanked members for giving their time to this issue.

Schiff framed the OPWG’s work in the context of how we think about pain within health care. Together we should be working to reduce suffering, and within that context, address opioid prescribing. The prescription opioid crisis is not one that state government can solve alone, and to that end Schiff and colleagues sought and obtained support for the Opioid Prescribing Improvement Program (of which this work group is a pivotal part) from health plans, health systems and health care providers.

Highlights of Schiff’s presentation include:

2

The US comprises just 4.6 percent of the world’s population, yet consumes 80 percent of theworld’s prescription opioids.

While Minnesota is faring better than most other states in terms of the prescription opioid crisis,there are harrowing disparities within the Native American population and to a lesser degreeamong non-Hispanic white Minnesotans.

DHS’ data show that the population of new, chronic opioid users is large and growing.Depending on the definition of new, chronic user, the number of such individuals is 5,000 to6,000 people a year. Eighty percent of new, chronic users have a recent history of behavioralhealth problems or chemical dependence or both.

Members discussed the large number of new chronic opioid users and disparities within the Native American population. In response to a question, Garrett clarified that DHS cannot tell what diagnoses are most prevalent among opioid claims. Pharmacy claims do not contain diagnostic information. To link a patient’s particular opioid claim to a particular diagnosis would require a chart review.

Tara Holt, who coordinates the state’s substance abuse strategy, reported on other state activities relevant to the OPWG’s work. In 2014 Minnesota was one of six states selected to participate in a National Governors Association-sponsored policy academy concerning prescription drug abuse. The academy led to the formation of a collaboration across state government called the State Opioid Oversight Project (SOOP). SOOP members comprise representatives of the departments of Human Services, Health, Education, Public Safety, Labor & Industry and Corrections, and the boards of Medical Practice, Nursing, Pharmacy, Dentistry, Veterinary Medicine and Podiatry.

The SOOP is focused on seven targeted focus areas:

Neonatal abstinence syndrome

Medication assisted treatment

Opioid prescribing

Prescription monitoring program

Increasing access to naloxone

Prevention

Increasing prescription take-back opportunities

In response to legislation passed in 2015, the state is implementing several of these strategies. The state is building a program to provide integrated care to pregnant women at risk of delivering a low-birth-weight baby or a baby who has been exposed to opioids in utero. The state is also implementing Steve’s law, designed to improve access to naloxone and distributing naloxone to emergency medical services providers via a one-time appropriation.

In August state and federal officials collaborated on a statewide summit called “Pain. Pill. Problem.” Over 1,000 people attended the event, which was at Northrop Auditorium at the University of Minnesota. Several community leaders presented at the event, including some members of this work group.

Ellie Garrett, who is currently serving as the OPWG’s primary staff support, summarized the legislation that authorizes the OPIP and calls for the OPWG’s creation (Minn. Stat. § 246B.0638). The OPIP will be a comprehensive project to improve opioid prescribing that is community-driven, and the community is represented by the OPWG. Specifically, the OPWG is charged with recommending:

3

Common protocols to treat:o Acute pain (0 – 4 days)o Post-acute (up to 45 days)o Chronic pain (> 45 days)

Sentinel quality improvement measures for each domain

Consistent messages for prescribers to give to patients

Criteria for identifying Minnesota Health Care Program providers that should undertake qualityimprovement

Criteria for terminating providers from MHCP whose prescribing practices are consistently andextremely outside of community norms

Garrett clarified that the OPIP does not apply to hospice patients or to treating cancer-caused pain. She shared a proposed work plan (attachment 1), that will likely be revised and become more detailed over time. While the OPWG is developing its recommendations, DHS will be working internally on the data reporting infrastructure that will be needed to implement the OPIP.

Schiff noted that OPIP’s quality improvement processes will be confidential—names of providers and their respective institutions will not be made public during the quality improvement phase. A member cautioned about the disruption patients face whenever a provider is terminated. Schiff stressed that DHS hopes that no providers would need to be terminated because of outlying opioid prescribing practices. Improvement should ensue from clearer community standards for opioid prescribing and consistent quality measurement.

Members discussed briefly various definitions of acute pain, including pain lasting the first X days from onset of painful condition or injury,

II. Opioids for Acute Pain

Garrett suggested that the work group consider endorsing or refining existing prescribing protocols rather than creating guidance from scratch. To that end, she shared three prescribing protocols that address treatments for acute pain: Washington state’s new guideline on prescribing opioids for pain (see Parts I, II and III on pages 9 – 31); ICSI’s protocol on acute pain assessment and opioid prescribing; and

recommendations from DHS’ Emergency Department Work Group (available from OPWG staff). She invited members to submit other guidance documents from their health systems or professional societies to inform discussions around acute pain.

Though each of these documents vary in specifics, they have in common a number of domains:

Assess and document function

Assess and document pain

Assess physical health contraindications/risks

Assess mental health and substance use contraindications, including family history

Consider take-home naloxone

Avoid concomitant use of benzodiazepines and other concerning drugs

Use lowest dose and duration

Specific guidance regarding type of drug (short/long acting, IM, IV)

4

Patient and family education about pain management, risks, benefits, reasonable expectations,safekeeping

Check PMP

Considerations regarding acute pain related/unrelated to concomitant chronic pain

Consider multi-modal, non-opioid treatments as alternatives or adjuncts to opioids

Provider/site-specific considerationso Dentalo Emergencyo Surgical

No public comments were offered.

Discussion among the members ensued. One member stressed the importance of access to addiction treatment, even though the OPIP’s emphasis is on prevention. Another acknowledged the large portion of medical board reviews pertaining to prescribing practices. When patient harm occurs, there are licensure consequences and a system in place to protect the public.

One member suggested that when the OPWG’s discussions turn to chronic pain, the group will have to address nuances of treating cancer patients, even though cancer-related pain is outside the OPIP’s scope. Sometimes cancer patients are still receiving pain medications begun during cancer treatment even after their cancer has gone into remission.

Another member observed that all too often what gets labeled by a patient or provider as acute pain is neither acute nor even pain. Other members agreed that diagnostic accuracy is critical.

A member suggested including a list of conditions, like migraine and uncomplicated neck and back pain where opioids are known to be ineffective. Those diagnostic exclusions should be contained in the protocol. Garrett acknowledged that ICSI’s guidance did address those exclusions, so the summary grid of domains being discussed should be expanded.

Members also discussed the 4-day duration in statute pertaining to acute pain. In contrast, Washington combines the acute and post-acute period in a single protocol.

After the break, Schiff opened the floor up again for public comment, but no comments were offered. Member discussion continued.

A member observed that risks of becoming a chronic user should be acknowledged when considering whether to initiate opioid therapy. People with low health literacy and pre-existing mental health conditions and/or pre-existing chemical dependency are at higher risk of chronically using opioids.

Another observed that ICD-10’s approach to severity indexing might help to better segment patients who need or do not need opioid treatment. Schiff expressed concerns with the tendency for diagnosis creep, which is a factor in billing; one hopes that it will not be a factor in opioid prescribing.

One member noted the importance of considering current incentive structures for clinicians—including incentives around lowering pain scores rather than treating pain appropriately. Other members agreed.

A member suggested that the group consider noting that opioids should not be considered as a first-line therapy. Another suggested that opioids should not be initiated without an endpoint and exit strategy,

5

both of which require a very clear diagnostic assessment of mental health, pain generators and substance use. He shared an anecdote about a young trauma patient who was struggling with mental health issues. Even in the emergency department, patients should always be screened for mental health or substance abuse, just as all patients who receive antibiotics are asked about allergies. Another member observed that ICSI’s ABCDPQRS protocol covers what providers need to know, so long as it is always used. Another member agreed with the importance of assessing mental and chemical health history even in trauma patients, given the propensity of some patients to self-injure in order to obtain opioids.

One member noted that checking the PMP goes a long way to identify drug-seeking behaviors, but access to the PMP is too restricted. Even when part of the treating team, pharmacists cannot access the PMP. They can do so only if they are filling the prescription. Some providers have addressed this access problem by requiring patients to consent to checking the PMP as part of admission. A member noted, though, that many drug-seekers will use multiple names to obtain medications, thus thwarting the PMP. A robust database alone in insufficient; provider education about drug seeking is also important.

One member reported another PMP problem, which concerns medication-assisted treatment for addiction. While suboxone dispensed at a pharmacy will be reported to the PMP, medications dispensed at a methadone clinic are not. In this regard, Minnesota’s privacy laws impede patient care. Similarly, Title 42 prohibits law enforcement personnel from monitory a drug treatment facility, even when criminal activity is suspected.

Schiff queried whether a requirement to check the PMP for every opioid prescription is reasonable. Is it necessary to check the PMP when treating a child with a long-bone fracture at a pediatric emergency room when there is no abuse suspected? It is important to not over-burden providers. Similarly, should a requirement to screen for mental and chemical health history be universal?

Members discussed varying perspectives on whether a naloxone prescription should be standard when initiating opioids. Some members suggested that patients whose personal, family or living circumstances put them at heightened risk of overdose or diversion should receive naloxone; others suggested that access to naloxone should be universal for all opioid recipients. Another member queried the relative priority for naloxone as a rescue treatment in relation to addiction prevention and treatment.

Members briefly discussed the importance of quality measurement and for providers to be able to see where their own prescribing patterns fall in relation to their peers. One member stressed the importance of normalizing the data with regard to patient risk adjustment, regional differences, provider types and other variables. Other members noted that other large systems are beginning to embracing measurement on prescribing patterns, such as the Veterans Administration and Medicare.

Schiff stated that the domains document would be revised and brought back for consideration at the December meeting.

III. Other business

Garrett distributed reimbursement forms to the members and asked them to watch for an email regarding scheduling for 2016 meetings. Future meetings will be supported by webcasts (for non-members) and full, two-way audio/visual connections for members.

The meeting was adjourned at approximately 3:30.

Attachment 1

6