Simeon D. Kimmel, Sophie Rosenmoss, Benjamin Bearnot, Zoe Weinstein, Shapei Yan, Alexander Y. Walley, and Marc R. Larochelle
Abstract
Referrals of hospitalized patients with opioid use disorder (OUD) to postacute medical care facilities are commonly rejected. We linked all electronic referrals from a Boston safety-net hospital in 2018 to clinical data and used multivariable logistic regression to examine the association between OUD diagnosis and rejection from postacute medical care. Hospitalized patients with OUD were referred to more facilities than patients without OUD (8.2 versus 6.6 per hospitalization), were rejected a greater proportion of the time (83.3 percent versus 65.5 percent), and in adjusted analyses had greater odds of rejection from postacute care (adjusted odds ratio, 2.2). In addition, people with OUD were referred disproportionately to a small subset of facilities with a higher likelihood of acceptance. Our findings document disparities in postacute care admissions for people with OUD. Efforts to ensure equitable access to medically necessary postacute medical care for people with OUD are needed.
Hospitalizations for people with opioid use disorder (OUD) in the US rapidly increased between 2006 and 2016, going from 164 to 297 hospitalizations per 100,000 people.1 Complications associated with opioid use, including systemic infections from drug injection, overdoses, physical and psychological traumas, strokes, or other acute conditions such as pneumonia and chronic obstructive lung disease, have contributed to the increase in hospitalizations.2–6 People with OUD commonly require prolonged intravenous antibiotics, wound care, medication titration, and physical or occupational therapy after stabilization from an acute hospitalization. For many, these services can only be delivered in postacute medical care facilities (for example, medical rehabilitation or skilled nursing settings).
Massachusetts has the second-highest rate of opioid-related hospitalizations in the US, making discharge planning and postacute care access for patients with OUD an especially important issue in the state, as these patients tend to have longer hospitalizations than patients without OUD with the same conditions.7,8 In 2016 the Massachusetts Department of Public Health issued guidance to all Massachusetts-licensed facilities that people with OUD should not be excluded from admission to postacute medical care because of treatment with medications for OUD (MOUD) such as methadone or buprenorphine.9 Despite this, the US Attorney’s Office for the District of Massachusetts has reached several settlements with postacute medical care facilities for violating the Americans with Disabilities Act of 1990 by screening out people with OUD or those treated with MOUD.10,11 Several clinicians have described the challenge of finding postacute care for people with OUD, but few studies have systematically evaluated postacute care referral and admissions practices.12,13 Previous work has shown that facilities frequently reject referrals explicitly because of substance use or MOUD, in violation of state and federal policies.10,14,15 In fact, in 2018 nearly four in ten patients with OUD referred for postacute care from a Massachusetts safety-net hospital were not able to be discharged to any postacute care facility.14 These rejections thus limit access to medically necessary postacute care and likely contribute to longer hospitalizations for people with OUD. However, it is not known whether people with OUD are more likely to be rejected from postacute care facilities when compared to those without OUD or whether they experience distinct postacute care referral patterns.
In this study we used data from Boston Medical Center’s electronic postacute care referral system to examine the association between OUD diagnosis and referrals to and rejection by postacute medical care facilities. We hypothesized that referrals for people with OUD would be more likely to be rejected than referrals for people without OUD and that people with OUD would be preferentially referred to a subset of postacute care facilities with a higher likelihood of accepting people with OUD, masking disparities in observed rejection rates. To test this hypothesis, we conducted a stratified analysis of acceptance rates by likelihood of facility to receive a referral for a patient with OUD.