Key takeaways:
• Hospitalized patients with OUD seen by a hospitalist-led consult service had much higher medication initiation rates than patients with OUD who weren’t referred.
• Hospitalists may not be comfortable prescribing OUD medications because of ongoing stigma or a lack of training.
• Patients benefit from strong support from social workers to maintain their treatment regimen as outpatients.
HOSPITALIZATION IS A KEY opportunity to initiate medications to treat opioid use disorder (OUD). Because Duke University Hospital, a quaternary center in Durham, N.C., does not have an addiction medicine service, hospitalists there in 2019 stepped up to launch a consult service to implement evidence-based treatment for patients with OUD.
That service now fields about 40 new consults a month. Called the Project for Caring for patients with Opioid Misuse through Evidence-based Treatment (COMET), the service just published its first four years of outcomes data in the Journal of General Internal Medicine.
According to that write-up, patients with OUD seen by the hospitalist-led service had much higher initiation rates of medications for OUD while hospitalized: 79% vs. only 42.7% for a concurrent group of hospitalized patients with OUD who were not seen by COMET (Most patients in the concurrent group likely had less acute medical needs than those referred to the service.) The percentage of patients receiving medications for OUD in the historic control group—those with OUD hospitalized at Duke in the three years before the service was launched—was 29.7%.
Patients seen by the service also had better inpatient and 30-day mortality rates as well as fewer readmissions. They did, however, have longer lengths of stay than concurrent controls.
“Some things should be standard practice for any hospitalist. Prescribing naloxone on discharge or distributing it is definitely one of those.”
Dana Clifton, MD
Duke University Hospital
According to lead author Dana Clifton, MD, a med-peds hospitalist who helped launch the service, that longer length of stay is likely due to clinicians making sure that patients are stable on their medication regimen and have a safe discharge plan. Part of a safe plan includes addressing this population’s high burden of unmet social needs.
Dr. Clifton notes that the study is one of the first to find evidence of a mortality benefit for patients with OUD seen by a hospitalist-led consult service. The service has also, she adds, “turned out to be a very rewarding part of my job as a hospitalist.”
While staffing the service “has been an unexpected turn in my career,” says Dr. Clifton, “it’s made me realize what a huge gap in care we had before the service existed.” While she knows most hospitalists are hesitant to begin treating OUD, “I want generalists to know that they can do this.” Dr. Clifton spoke to Today’s Hospitalist.
It’s well known that patients with OUD are frequently hospitalized and that medication-assisted treatment saves lives. So why do you think so few hospitalists initiate treatment?
I think it is primarily a lack of comfort and training in using medications for OUD. That goes back to medical schools and residency programs needing an increased focus on addiction medicine, although that landscape is starting to change.
I also think that both health systems and providers have ongoing stigma based on the false assumption that addiction is due to a moral failure or a lack of willpower. Some clinicians still don’t see opioid use disorder as they should: as a chronic disease like hypertension and diabetes.
And stigma prevents health systems from creating policies and guidelines to help providers, so they end up being less interested in caring for this population. But once you start caring for patients with OUD, you realize how rewarding the work can be.
Since you established this service, are more of your hospitalist colleagues who aren’t on the service more comfortable prescribing these medications themselves?
They do have a greater comfort level managing patients with OUD, and that’s not just physicians. Nurses, ancillary staff, pharmacists—they are all more comfortable with and knowledgeable about these medications.
It always warms my heart when hospitalist admitters or residents who admit patients recognize and treat opioid withdrawal by starting buprenorphine or methadone. We still get consulted on those patients to help titrate the medication and support patients’ post-discharge transition. But the service has changed our institution’s culture in a positive way.
How many hospitalists rotate through the service and on what type of schedule?
We have a pretty stable group of about 15 hospitalists. Some have left to pursue other interests, and new hospitalist providers have expressed interest in rounding on COMET, though we are currently fully staffed. A lot of us have been on the service for the full six years, so our retention rate is high.
Everyone on the service has to do a minimum of 10 consult shifts every six months to maintain competency. The shifts count of course toward each doctor’s annual clinical complement. We try to do at least a few days in a row to maintain continuity with patients— ideally closer to five days in a row.
When we launched the service, we were able to hire one dedicated social worker to help secure outpatient resources for patients and address their unmet social needs. We have since been able to hire an additional social worker, so the service now has two full-time.
What case did you make to your administration to green-light the service?
When we first brought up the idea more than eight years ago, our main argument was that we weren’t doing the right thing for these patients by not treating their opioid use disorder. Patients were being admitted with endocarditis or complex infections, which we treated with antibiotics. But we weren’t treating their OUD, which is the underlying disease process.
We also initially made the case that such a service might decrease patients’ length of stay. As we note in our paper, that didn’t happen. We thought we’d be able to partner with SNFs to get some patients who needed prolonged IV antibiotics out of the hospital sooner. But then covid hit, and that didn’t materialize.
We believe the improved quality of care that COMET provides with reduced mortality and readmission rates offset patients’ slightly longer length of stay.
But six years in, you have no pushback from administration?
None. We have full support for this service.
Your study mentions how the mortality benefit from your service wanes over 90 days post-discharge. That speaks to high recidivism and the fact that patients in this population are often lost to outpatient care. Have you advocated for more robust outpatient resources?
We have. Part of that is our increased social work support to try to address housing, transportation and other barriers to getting outpatient treatment.
We’ve also advocated to expand the number of primary care providers who prescribe buprenorphine. It’s challenging because we are a big referral center, so close to half our patients aren’t from our county or adjacent counties. Some of our patients can’t follow up within our own health system, which is why we need such robust social work support to connect patients to OUD care in their local communities.
The study mentions that COMET implemented a methadone protocol in 2020. You write that 72.6% of patients on your service received buprenorphine while another 19.8% were administered methadone. You also mention a growing need for methadone in the fentanyl era. Why is that?
We have heard anecdotally from patients that methadone is more effective in treating opioid cravings and opioid withdrawal. A Canadian study published in JAMA in 2024 supported what patients tell us—and demonstrated that patients with OUD on methadone continue in treatment longer than those on buprenorphine.
Hospitalists are more comfortable prescribing buprenorphine than methadone, in part because outpatient follow-up for buprenorphine for OUD is much easier. Any provider—including those in primary care—can prescribe it. Patients on methadone must go to an outpatient treatment program with often daily dosing, at least initially.
Methadone also has more drug-drug interactions and can affect the QTc interval, which can pose additional complexity for patients with multiple medical issues who are on a lot of medications. Still, we wanted to demonstrate that hospitalists can prescribe methadone to inpatients with OUD or opioid withdrawal. While we can’t prescribe it at discharge, hospitalists can start methadone in the hospital and connect patients to an outpatient treatment program to continue after discharge.
As for fentanyl, it’s more potent than heroin. We needed updated national policies and guidance reflecting that change to better treat OUD and opioid withdrawal.
Even in facilities where hospitalists don’t prescribe these medications, what should doctors do to mitigate patients’ risk?
Some things should be standard practice for any hospitalist. Prescribing naloxone on discharge or distributing it is definitely one of those. Even though it’s now sold over the counter, the cost can be prohibitive, so prescribing or distributing it removes that barrier.
Hospitalists should also know and tell patients about harm-reduction organizations in their area. Even if hospitalists can’t distribute items like fentanyl and xylazine test strips, it’s helpful to tell patients where they can get those and safe-injection supplies.
And I think hospitalists have a role in educating these patients in harm reduction. Although many patients with opioid use disorder could educate us on safe injection practices and overdose prevention, it still helps to talk patients through those to make sure they understand and utilize them. We discuss how they inject and what their risks of infection are from non-sterile or risky injection practices. We also talk in detail about overdose prevention, including not using alone or using the Never Use Alone hotline.
Hospitalists should also talk to patients about decreased tolerance after hospitalization and using test doses if patients are obtaining drugs from a different drug supply.
Since you launched this service, you’ve started taking consults for hospitalized patients who are pregnant. Any plans to expand the service further?
I’m always thinking about how to improve our service. We are currently exploring how to expand to other hospital sites within Duke Health, to find and train champions at those sites and figure out how to replicate our policies and protocols.
Because transitioning to outpatient care is such a challenge, we’d love to create a bridge clinic and have a group of outpatient provider partners to whom we can refer. I also envision some type of e-consult or virtual model where we could talk other providers through a challenging case—or even a straightforward one that may be challenging to others. That way, providers could become more comfortable initiating care on their own.
Phyllis Maguire has been Executive Editor of Today’s Hospitalist since 2006. Based in Bucks County, Pa., her health care interests are hospital medicine and long-term care options. She also likes zydeco, hiking, and reading memoirs and romances.
A study in Annals of Internal Medicine found that patients receiving methadone in primary care over two years also accessed many more health care services, including screenings for cancer, hepatitis and HIV.
