OBJECTIVES/GOALS: Medications to treat opioid use disorder (mOUD) are available and can save lives, but are underutilized. We hypothesize that the rate of prescribing varies by treatment facility and these differences will shed light on barriers and facilitators to mOUD utilization. METHODS/STUDY POPULATION: We performed an exploratory analysis in MD Clone, a platform which generates non-identifiable synthesized data based on real patient data in the electronic health record (EHR) of St. Louis based hospitals. Our query included adults aged 18-70 with an OUD diagnosis using ICD-9 of -10 codes (opioid abuse, opioid dependence, opioid poisoning, opioid withdrawal) occurring between 2013 and 2022 along with prescriptions for buprenorphine, methadone, or naloxone within 7 days of the condition being entered in the record. We compared the rate of medication prescription within 7 days across settings and facilities where the patients were seen. We propose to replicate this analysis in actual patient records from the EHR following IRB approval. RESULTS/ANTICIPATED RESULTS: Our synthetic data comprised 24600 patient diagnoses. After filtering for patients seen in the ER or inpatient 16235 patients remained in the data set. Of these, 4376 fell into one of the categories that clearly warrant treatment with medication. Out of 4376 patients with a qualifying OUD related condition, only 815 (18.6%) received a prescription for any of the medications. Rates of prescribing within facilities varied between 67.2% of eligible patients receiving a prescription at a rural location to 0% at some urban centers. We anticipate similar findings from analysis of patient records obtained from the EHR. We will extend our analysis to explore factors which may be driving the wide difference in prescribing to better understand barriers and facilitators to mOUD utilization. DISCUSSION/SIGNIFICANCE: We identify under-utilization with differences across facilities in prescribing mOUD based on preliminary work in synthetic data. If true, this represents a gap in care and opportunity for intervention. By replicating the MD Clone results in patient data from the EHR we will confirm this finding and increase acceptability to clinicians.