Alex Walley and colleagues have published their analysis of the Massachusetts naloxone program, which shows what I'm going to call a per capita dose-dependent reduction in opioid overdose death (rate ratio of 0.73 [95%CI 0.57-0.91] if naloxone was given to 1-100 people per 100,000 population and 0.54 [0.39-0.76] for naloxone to >100/100,000). This provides the first real community-level effectiveness data to-date.
The finding that naloxone was more effective the more people per population were given kits suggests that aggressive distribution may be the optimal approach for naloxone programs. This study doesn't answer the question as to why there was a dose-dependent response - was it simply because more is better? were there social network effects where naloxone became substantially more effective if it reached into certain social networks, which was more likely if it was more widely distributed? Social network issues have yet to be explored in overdose prevention research.