Nudging Emergency Department-Initiated Addiction Treatment.

Abstract

When initiated in the Emergency Department (ED), medication for addiction treatment (MAT) with buprenorphine improves outcomes, increases engagement in addiction treatment and decreases the use of inpatient addiction treatment services. Unfortunately, initiating MAT in the ED is not yet standard practice. We assessed the impact of the addition of a multipart behavioral science-based intervention to increase opioid use disorder (OUD)-related treatments prescribed in the ED. Our ED initiated a campaign to help ED faculty obtain their DEA-X waiver required to prescribe buprenorphine. In parallel, we implemented 2 ED-initiated buprenorphine treatment pathways. We then conducted a two-stage qualitative process informed by behavioral science to identify key barriers to physician use of the MAT protocol. Using these insights, we developed 4 behavioral science-based interventions. To assess the impact of the interventions on the number of OUD-related treatments per day among patients meeting the inclusion criteria we compared the number of OUD-related treatments per day before versus after the interventions began using t tests. Then, in our primary model, we estimated the causal effect of the behavioral interventions using a regression discontinuity in time approach. Across the entire year study period, there is an increase in OUD-related treatment after the interventions begin, driven by greater use of ambulatory referral orders. The unadjusted mean difference in any OUD treatments per day pre- versus post-intervention increased by 0.80 (95% confidence interval [CI]: 0.04, 1.56; P = 0.039) whereas the number of ambulatory referral orders placed increased by 0.82 (95% CI: 0.48,1.16; P < 0.001). Using the 120-day study window and an ordinary least squares regression discontinuity in time model, the 4-part intervention increased the number of patients receiving any opioid treatment in the ED by 1.6 additional treatments per day (95% CI: 0.04, 3.19; P = 0.045). To support our protocol and increase the provision of ED-MAT, we implemented 1 patient-facing and 3 provider-facing interventions rooted in behavioral science principles. Our results show that this pack of behavioral science interventions increased the likelihood that ED providers offer MAT to patients with OUD.