Abstract

Severe opioid withdrawal, risk of patient-initiated discharge, and some inpatients’ use of unregulated substances prompt clinical and ethical questions considered in this commentary on a case. Short-acting opioids can be used to manage inpatients’ pain and opioid use disorder (OUD) withdrawal symptoms. Including evidence-based interventions—such as naloxone kits, substance use equipment, and supervised consumption—in some inpatients’ care plans may make those patients safer and reduce their risk of death. These and other strategies align with clinicians’ ethical duties to minimize harms and maximize benefits for inpatients with OUD.

Case

KC is admitted for infective endocarditis secondary to microbes entering their bloodstream during repeated injection drug use. KC has a long history of opioid use disorder (OUD) and has intermittently been treated for it. KC’s history of using opioids started when KC exhausted a supply of oxycodone, prescribed with limited refills for postoperative pain management, which led KC to start using heroin and then fentanyl.

Members of KC’s clinical team have not come to consensus about how to manage KC’s pain or OUD. They are aware that KC has their own supply of drugs and wants to leave the hospital as soon as possible and against medical advice, if necessary. Team members consider administering short-acting opioids to keep KC comfortable and in hospital for intravenous antibiotics and evaluation for cardiac surgery, but one clinician opposes any care plan that “feeds” KC’s OUD.

Commentary

People who initiate their own discharge from hospital have a well-documented increased risk of death,^1,2 and people who use substances are at greater risk of premature discharge than other groups.³ Every effort must therefore be made to engage people in care that offers concomitant management of their primary medical condition and any substance-related diagnosis. For people with OUD, such care includes immediate access to all forms of opioid agonist treatment (OAT)^4,5 and effective management of pain and withdrawal. OAT options include buprenorphine formulations, methadone, slow-release oral morphine, and injectable treatments such as hydromorphone and diacetylmorphine.⁴ Of note, slow-release oral morphine and injectable treatments are not currently available in the United States for people with OUD.

Even if they offer these interventions, hospitals must accept that not all individuals will stop using regulated or unregulated substances. To reduce morbidity and mortality risk from ongoing substance use while in hospital, access during hospitalization to other interventions such as naloxone kits, clean substance use equipment (eg, syringes, cookers, sterile water), and supervised consumption services should be considered.

Prioritizing Harms for Reduction

Although the standard of care for the treatment of infectious endocarditis is several weeks of intravenous antibiotic therapy, a scoping review found that published guidelines on the management of endocarditis in people who inject drugs rarely recommend addiction medicine consultation or opioid agonist treatment, and none discuss withdrawal management.⁶ These guidelines suggest that OUD is considered unique or separate from other medical needs during a hospital admission, supporting the perspective that treatments and referrals that respond to a patient’s substance use are optional or exceptional. In many cases, this belief is an error. The method for assessing which harms should be prioritized for a patient with OUD should be similar to, if not the same as, the method used for a patient with multiple comorbidities that do not include OUD.

Clinical judgments about the potential likelihood and severity of harms are the first important step in determining which clinical needs should be given priority. KC has 2 urgent medical concerns: infectious endocarditis and OUD. Discussions with KC should include how treatment of one clinical need might be necessary to facilitate treatment of the other; in this case, offering OAT⁷ and short-acting opioids for withdrawal management is a means of enabling access to a complete course of antibiotics for endocarditis.

KC’s goals, values, and beliefs should inform which clinical needs take priority. KC might not wish to initiate long-term treatment for their OUD during this admission, but they might be interested in other interventions that could reduce their risk of negative health outcomes related to substance use and help them achieve other important goals. As with any patient, KC ought to be given choices regarding treatment options (including the option not to treat), and those choices should be respected, even if they do not align with what the care team sees as optimal or most appropriate.

Risks and Benefits of Short-Acting Opioids

People with OUD who regularly use illegally manufactured synthetic opioids (eg, fentanyl, carfentanil) are likely to have developed a high tolerance to opioids.⁸ As such, opioid withdrawal should be anticipated. While non-opioid medications such as acetaminophen, ibuprofen, clonidine, and others can help with the symptomatic management of some opioid withdrawal symptoms, they will not meet the baseline opioid requirements or severe acute pain and withdrawal management needs of individuals who regularly use highly potent synthetic opioids.⁹ Short-acting opioids can be titrated to effectively manage acute pain and withdrawal in patients who are started on OAT, such as buprenorphine-naloxone or methadone or slow-release oral morphine titrated over days to weeks; one meta-analysis found that the rate of all-cause mortality during OAT is more than half the rate during time out of OAT.¹⁰ Short-acting opioids can also be used to manage acute pain and withdrawal in hospitalized patients who decline OAT initiation.^9,11

Undertreated withdrawal and pain are two of the main drivers of patient-initiated discharge in hospitalized patients with OUD.¹² The risks of patient-initiated discharge for patients like KC include complications related to untreated infective endocarditis such as heart valve dysfunction, septic emboli, worsening systemic infection, and death.¹³ They also include the risks of untreated active, severe OUD, such as drug poisoning death, traumatic injury, suicide, and complications related to injection drug use (eg, blood-borne infections like HIV or hepatitis C).¹⁴ In addition to reducing these risks, hospitalization can offer other benefits, such as screening and treatment for sexually transmitted and blood-borne infections; vaccinations; assistance with housing and income support applications; and referral to community-based primary care, addiction treatment, and other services.

Are there risks to KC, or to society in general, if short-acting opioids are used in a hospital setting? KC is already physically dependent on high-potency synthetic opioids and likely meets the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria for severe OUD.¹⁵ It is unlikely that the short-term use of less potent, regulated opioids in a hospital setting will worsen the severity of their OUD. People with severe OUD, by definition, will continue to use opioids despite ongoing negative health and social consequences. While concern exists over the diversion of prescription opioids into the community at large, this concern can be largely mitigated in hospitals by ensuring that short-acting opioids are prescribed in formulations that are harder to divert (eg, liquid) and that ingestion is witnessed by a regulated health professional. While the risks to KC or society in general are likely small in the hospital setting, unique legislative restrictions on the prescribing and dispensing of opioid medications in the United States (Administering or Dispensing of Narcotic Drugs¹⁶) and Alberta, Canada (the Mental Health Services Protection Act¹⁷ and Mental Health Services Protection Regulation¹⁸) can contribute to confusion about where, when, and for how long medications like short-acting opioids can be used, which in turn potentially contributes to inaction and undertreatment of pain and withdrawal.

Both hospital- and patient-initiated premature discharge can preclude patients from accessing high-quality medical care. 

As part of the informed consent process, the duration for which short-acting opioids will be prescribed should be discussed in advance with the patient. There is considerable variation across North America in the use of short-acting opioids for people with OUD in community-based settings.^19,20 Whether these medications will be tapered prior to or after discharge or continued with the transfer of care to another prescriber in the community should be discussed prior to their initiation in hospital.

Managing Risks of Nonprescribed Substance Use

Even with expertly managed pain and withdrawal and access to the full continuum of treatment options for OUD (both of which should be part of the standard of care in hospital settings), some people might continue to use their own substances while admitted.²¹

The ongoing use of nonprescribed substances in hospitals might present risks to the patient (eg, unattended drug poisoning event, recurrent infections from lack of access to clean supplies) and to staff (eg, contact with used injection equipment). Abstinence-oriented hospital policies might also place staff and patients in conflict when sanctions for ongoing substance use are implemented (eg, hospital-initiated discharge, constant patient surveillance, revocation of off-unit privileges).²² Both hospital- and patient-initiated premature discharge can preclude patients from accessing high-quality medical care.²³

These risks can be mitigated by taking a more pragmatic, harm reduction-oriented approach to care,^22,24 which can include the integration of interventions that have been well studied in community settings into hospital-based care. Naloxone kit distribution, safer substance use education for patients and staff, access to new consumption equipment and sharps disposal containers, and ensuring patients have access to secure storage might all help reduce the risks associated with substance use in hospitals. Access to supervised consumption services that provide sites where hospital patients can consume their own substances under medical supervision is another example of how hospitals have tried to reduce the risks to patients of taking a nonprescribed substance.^25,26 Formalized supervised consumption services offer several advantages over other ad hoc measures, such as protection from illegal drug possession charges for staff and patients while following the approved policies and procedures of the service, safe disposal of used equipment, and the availability of an immediate medical response to any adverse reactions.

Language Use When Caring for People With OUD

Patients who use substances often experience being stigmatized and mistreated by health care professionals during hospital admissions.²⁷ The language used to describe patients with OUD affects how health care professionals (and others) judge and value these patients, perceive the cause of the problem, and view whether the patient is deserving of treatment.²⁸ Stigmatizing language (eg, substance abuser, addict) should be strenuously avoided in favor of more neutral language that recognizes patient dignity and emphasizes a medical approach to OUD.

In cases like KC’s, questions have been raised about whether there is value in using terms like life-threatening to describe risks associated with failing to offer resources to address patients’ OUD. In general, any terminology that is used should be accurate and, when possible, supported by evidence. If a course of action or lack of action is life-threatening, then it should be described as such. OUD would certainly be considered a life-threatening or life-limiting diagnosis, with the average life expectancy of people who have been prescribed OAT being approximately 15 years shorter than that of the general population.²⁹

That said, it is not clear that using terms like life-threatening will motivate health care practitioners who hold stigmatized views of OUD to take steps to address a patient’s OUD. If OUD is perceived as a character flaw or a concatenation of “bad” individual decisions (which might be the perception of KC’s clinician, who is concerned about “feeding” their OUD), then describing OUD as life-threatening might not result in someone seeing a greater need to act. Death might simply be seen by some clinicians as the unfortunate result of an individual’s poor choices—with no burden of responsibility for medical professionals or the health care system, politicians, or society to bear.

This view that no one bears responsibility for the preventable (though perhaps regrettable) death of a patient with OUD because it is the result of that patient’s choices stands in stark contrast to general societal expectations of the role of physicians and hospitals. We expect—and demand—that hospitals provide care for the most urgent medical, surgical, and psychiatric issues 24 hours a day, 365 days per year. There is an expectation that all efforts will be made to provide life-preserving care, whether or not the life-threatening circumstances experienced by patients were the result of their own actions or choices. Hospitals offer the most advanced and intensive treatment options for (almost) all medical conditions, yet, in most hospitals in North America, access to physicians with addiction medicine expertise is rare,^30,31 and access to specialized treatment options (such as injectable OAT like diacetylmorphine or hydromorphone³²) is virtually nonexistent. This lack of access to OUD care in hospitals is not only inconsistent with the broader expectations of health care, but also discriminatory and in violation of the Americans with Disabilities Act.³³ When faced with emerging threats, such as COVID-19, the health system was able to respond quickly to deliver new expertise, testing, medications, and vaccinations. Yet interventions proven to reduce deaths in people with OUD have not been spread and scaled in a commensurate way.

Conclusion

People with OUD in hospital settings urgently require access to a full continuum of evidence-based OUD care that is provided without stigma or judgment. Such care is consistent with clinicians’ ethical duties to minimize harms and maximize benefits for their patients and to set the conditions whereby patients might optimally benefit from treatment of their acute medical illness. Access to all forms of OAT (buprenorphine, methadone, slow-release oral morphine, and injectable formulations), naloxone kits, clean substance use equipment, safe disposal of used equipment, and supervised consumption services in combination have the potential to dramatically decrease the risks that patients with OUD currently experience in hospital settings. Providing patients evidence-based care for both their OUD and their other conditions, without negative bias, is the standard generally expected for treating any condition.