Development and evaluation of a pilot overdose education and naloxone distribution program for hospitalized general medical patients.

Overdose education and naloxone distribution (OEND) to people at risk of witnessing or experiencing an opioid overdose has traditionally been provided through harm reduction agencies. Expanding OEND to inpatient general medical settings may reach at-risk individuals who do not access harm reduction services and have not been trained. An OEND program targeting inpatients was developed, piloted, and evaluated on 2 general medicine floors at Montefiore Medical Center, a large urban academic medical center in Bronx, New York. The planning committee consisted of 10 resident physicians and 2 faculty mentors. A consult service model was piloted, whereby the primary inpatient care team paged the consult team (consisting of rotating members from the planning committee) for any newly admitted patient who had used any opioid in the year prior to admission. Consult team members assessed patients for eligibility and provided OEND to eligible patients through a short video training. Upon completion, patients received a take-home naloxone kit. To evaluate the program, a retrospective chart review over the first year (April 2016 to March 2017) of the pilot was conducted. Overall, consults on 80 patients were received. Of these, 74 were eligible and the consult team successfully trained 50 (68%). Current opioid analgesic use of ≥50 morphine milligram equivalents daily was the most common eligibility criterion met (38%). Twenty-four percent of patients were admitted for an opioid-related adverse event, the most common being opioid overdose (9%), then opioid withdrawal (8%), skin complication related to injecting (5%), and opioid intoxication (2%). Twenty-five percent had experienced an overdose, 35% had witnessed an overdose in their lifetime, and 83% had never received OEND previously. Integrating OEND into general inpatient medical care is possible and can reach high-risk patients who have not received OEND previously. Future research should identify the optimal way of implementing this service.