Center for Opioid Recovery + Engagement
2019 • Penn Medicine Center for Health Care Innovation
Problem + opportunity
Philadelphia has the highest rate of opioid overdose death of any major city. There are about 2,100 yearly visits from patients with opioid use disorder (OUD), at Penn’s three central emergency departments (ED).
Treating OUD isn’t easy, but there are tools that work. Medication-assisted treatment (MAT) is 60-80% effective. Certified recovery specialists (CRS) connect patients to treatment and support them during recovery. CRS’s are peers, meaning they are also in recovery.
Before this work began, these tools were under-used. OUD patients were typically stabilized and then discharged. OUD is a medical condition, but Penn’s ED treatment failed to address it. I joined a team of providers and administrators to fully incorporate these evidence-based tools into ED care.
Action
Identifying barriers
MAT and CRS support were available in the ED, but were under-used. Using contextual inquiry, interviews, and data analysis, we identified key barriers.
Enhanced treatment
CRS’s relied on providers to request a consult when OUD patients arrived in the ED - a manual process that was easy to forget. We hypothesized that an automated system could proactively identify OUD patients and notify the CRS’s in a timely way.
I led a pilot which tested this hypothesis using manual effort. Not only did we identify more patients, the CRS’s were able to take meaningful action. Using learning from this pilot, we developed a machine-learning algorithm to identify OUD patients and alert our CRS’s via secure text.
ED treatment
Once patients were identified, we wanted to ensure they were offered evidence-based treatment, with an emphasis on buprenorphine (bupe). When we started, few ED doctors were familiar with bupe. Using financial incentives, we nudged them to complete targeted training. Now, over 90% of ED faculty are prepared to use bupe.
Finally, to encourage and reinforce the use of bupe as a first-line treatment, we used social norming and behavior change strategies. For example, when providers used bupe for the first time, we public thanked them and provided them with a pin. Each Friday we circulated success stories from the week to build excitement around substance use treatment.
Impact
CORE is a growing program. CORE is now an established, growing program serving all three of Penn’s central hospitals.
More patients are receiving peer support. CRS’s went from consulting 4 OUD patients per month to over 100. More patients are starting treatment. We went from starting 20% of eligible ED patients on bupe to 68%.
Patients are staying in treatment. 68% of patients consulted by a recovery specialist and started on bupe remained in treatment for at least 30 days, compared to <5% at baseline.
We’re preventing overdoses. We estimate that our work will prevent 200 overdoses per year.
Articles
CORE overview
Law 2030 presentation. Presentation starts at 2:36:00.
