Trends and characteristics during 17 years of naloxone distribution and administration through an overdose prevention program in Pittsburgh, Pennsylvania.

OBJECTIVE: Describe time trends during 17.5 years of community-based naloxone distribution. METHODS: Analysis of administrative records from a harm reduction program in Pittsburgh, Pennsylvania, USA, collected during encounters for overdose education, naloxone dispensing and refills. Monthly time trends were analyzed using segmented regression. Programmatic context aided interpretation of quantitative findings. We also evaluated impacts of 2014 state legislation loosening naloxone prescribing requirements and providing Good Samaritan protections. RESULTS: From July 2005 to January 2023 there were 16,904 service encounters by 7,582 unique participants, resulting in 70,234 naloxone doses dispensed, with 5,521 overdose response events (OREs), utilizing 8,756 naloxone doses. After legislation, new participants increased from 10.4 to 65.9 per month. New participants tended to be older (46 vs. 37 years), female (58% to 35%), White race, and more likely to be family/friends as opposed to people who use drugs themselves. Consequently, ORE per participant fell from 1.46 to 0.47 in the year after enactment. On average, 1.63 (95% CI: 1.60, 1.65) naloxone doses were administered per ORE, which did not change substantially over 17 years (χ2 = 0.28, 3 df, p = 0.60) during evolution from prescription opioids, to heroin, to illicitly manufactured fentanyl. In 98.0% of OREs the person who experienced overdose "was okay", i.e., survived. Emergency medical services were called in 16% of OREs overall, but <7% since 2019. There were 106 more emesis events per 1,000 OREs with 4 mg nasal spray compared to intramuscular injection; and 48 per 1,000 more reports of anger. Titration of intramuscular naloxone was associated with lower rates of adverse events. CONCLUSIONS: While state legislation created the environment for expansion, reaching previously underserved communities required intentional new programmatic development and outreach. Long-term consistency of <2 doses per ORE, high survival rate, and robust utilization all lend confidence in prioritizing naloxone distribution directly to people who use drugs and their social networks. Trial registration: This investigation was pre-registered https://osf.io/b2f4h.