Viewpoint January 31, 2022: Short-Acting Opioids for Hospitalized Patients With Opioid Use Disorder

Afew months ago, I found one of my patients panhandling on his usual corner. He was in bad shape. Maggots were crawling through a wound where he regularly injected drugs and his bandages were soaked with pus. He was desperate to return to the hospital. First, though, he needed money for a “stash,” a supply of opioids to use while hospitalized. After leaving the hospital half a dozen times because of untreated pain and withdrawal, he had adapted. Now he always came prepared, even if it meant delaying care for days.

Overdoses claimed more than 93 000 lives in the US in 2020, more than in any prior year. Amid this ongoing crisis that is now fueled by synthetic opioids like fentanyl, individuals with opioid use disorder (OUD) too often see hospitals as places to avoid rather than as places of healing.¹ Countless individuals delay care despite serious illness.² Once they do present, nearly half will at some point use illicit drugs while hospitalized.³ Compared with those without addiction, hospitalized patients with substance use disorders are 3 times more likely to leave before they are medically stable, a decision associated with a doubling of all-cause mortality.⁴

Ask those with OUD why they delay care, use while hospitalized, or leave early and they consistently point to untreated pain, withdrawal, and opioid craving.^1–5 Given the severe consequences of these 3 symptoms, it may be time to consider a different approach: we could offer short-acting opioids to hospitalized patients with OUD. Although it may seem radical to some, this approach could quickly and effectively alleviate pain, withdrawal, and opioid craving, thereby facilitating treatment of the medical and surgical complications of addiction and of OUD itself.

Guidelines from the American Society of Addiction Medicine recommend nonopioids, buprenorphine, or methadone for withdrawal and pain in patients with untreated OUD.⁶ Nonopioids are helpful adjuvants, but when used without opioids, they increase the risk of overdose after discharge because patients lose tolerance.⁶ Buprenorphine and methadone (along with extended-release naltrexone) are effective for long-term addiction treatment and should be offered to all hospitalized patients with OUD.⁶ However, they are sometimes insufficient for initial symptom management. Buprenorphine increasingly risks precipitating withdrawal as illicit fentanyl replaces heroin across North America; some patients have responded by delaying or declining treatment with buprenorphine.⁷ Meanwhile, methadone takes 4 hours to reach a peak effect and more than a week of daily dosing to achieve a steady state that is therapeutic for withdrawal and craving.⁶ Without alternatives, many patients will treat themselves covertly with a contaminated supply of opioids. Others will leave early to do so.

Instead, we could administer short-acting opioids ourselves. These could be used in 3 sometimes overlapping scenarios, the first of which is to treat acute pain. Patients with OUD have high opioid tolerance and require higher doses of short-acting opioids for effective analgesia; treatment with buprenorphine or methadone dosed solely for OUD is rarely sufficient.^6,8 Second, short-acting opioids can treat withdrawal and craving while initiating treatment with methadone or can bridge patients to treatment with buprenorphine while illicit fentanyl clears. Last, short-acting opioids can treat withdrawal and craving in patients who are offered but decline treatment with methadone and buprenorphine. This last scenario, admittedly the most controversial, acknowledges that some patients are unable or unwilling to stop using short-acting opioids, even when offered treatment. We can accept this, offer a safer alternative to illicit opioids, and still treat the conditions that require hospitalization. Used in these scenarios, short-acting opioids could mitigate some of the most severe consequences of OUD by allowing patients to engage in treatment for conditions such as endocarditis and bacteremia.

Addiction consult services in Canada and the US have started demonstrating how this approach can be implemented.⁹ Short-acting opioids should be administered at a higher dosage for patients with OUD than for opioid-naive patients. Doses could be increased stepwise until patients report relief from pain and craving and show no signs of withdrawal. To reduce the risk of iatrogenic overdose, administrations should be spaced to avoid dose-stacking. For patients amenable to long-term treatment or being discharged to nursing facilities, short-acting opioids can be transitioned to methadone or buprenorphine; for those returning to illicit opioid use, they could be abruptly discontinued before discharge. In the US, according to Title 21 CFR §1306.07, there is no legal restriction to administering opioids in this manner as long as they are given as an incidental adjunct to medical or surgical treatment.

Given their addictive risk, this is not a call to be cavalier with opioids in general. But for patients who already have active OUD, the risk of developing addiction is no longer germane. Still, there is a risk that short-acting opioids could reinforce addiction. Does this risk eclipse the benefits of retaining patients in care? There is little evidence among hospitalized patients to determine the relative trade-offs. However, among outpatients, the evidence is fairly robust: 8 countries offer injectable opioid agonist treatment based on 7 randomized clinical trials demonstrating that patients with OUD that was refractory to treatment with methadone or buprenorphine had improvements in physical health, mental health, and retention in care when offered injectable diacetylmorphine (heroin) or hydromorphone.¹⁰ Thus, prima facie, for patients with active OUD who require hospitalization for life-threatening or limb-threatening infections, it seems unlikely that the marginal reinforcement from administering short-acting opioids for a few days would outweigh the benefits of continued care.

I recently used short-acting opioids in this manner with a young woman who was admitted for an injection-related arm wound. She had been using about $300 of illicit fentanyl daily. Initial treatment with buprenorphine had precipitated withdrawal, prompting her to self-administer illicit fentanyl (she also brought a stash). When I saw her as part of my institution’s addiction consult service in Baltimore, she was experiencing severe withdrawal despite receiving a dose of methadone. Initially, we used escalating doses of intravenous hydromorphone to treat her symptoms. Over the next few days, we increased methadone doses to therapeutic levels, scheduled oral hydromorphone administrations, and continued intravenous hydromorphone as needed for pain, withdrawal, or opioid craving. Her hand was eventually amputated because of osteomyelitis with a nonsalvageable radioulnar joint. By the end of a 12-day hospitalization, she was weaned off of hydromorphone and discharged to an opioid treatment program to continue treatment with methadone.

Tragically, this was not the first time she had sought care for her wound. She had been hospitalized for the same infection 3 months earlier. At that time, she was administered antibiotics, methadone, and low doses of oral hydromorphone, but her opioid doses were never escalated. After 36 hours of uncontrolled pain and opioid craving, she left the hospital. It was 3 months, 3 months with a progressive, untreated infection, before she returned. Had she received adequate doses of opioids during that first admission, could her hand have been saved?

We need further research to better understand how and when to implement this approach, especially in hospitals without addiction consult services. Additionally, even if used optimally, opioids and medications for treating OUD are only part of the story. Hospitalized patients with OUD also face stigma, discrimination, and barriers to accessing evidence-based harm reduction services.⁵ Nevertheless, as we enter the third decade of a worsening overdose crisis and face an increasingly contaminated supply of illicit opioids, it may be time to consider that adequate doses of short-acting opioids can serve as one component of compassionate, effective care of hospitalized patients with OUD.