The effects of syringe service programs on overdose deaths in Tennessee

Background
Opioid use, including heroin, prescription pain killers, and most recently fentanyl has become a public health crisis throughout the United States. In 2020 alone, more than 68,000 death were attributed to opioid use, nearly nine times as many deaths as in 1999 and a 46% increase from 2018. Harm reduction is a method of drug use interventions focused on reducing the risks associated with illicit drug use including overdose and infectious disease. One such intervention is syringe services programs (SSP), also known as needle exchanges. Initially designed to prevent the spread of infectious disease, SSPs have recently begun to be hubs of harm reduction services and oftentimes distribute naloxone, fentanyl test strips and other forms of drug testing. 
Methods
Data from all 95 counties in Tennessee are used including demographics, fatal overdose, nonfatal overdose, as well as if the county had a legal SSP in a given year. Counties were treated as having an SSP if an organization within the county applied for and was granted licensure to operate within a given year. Five of the ninety-five counties were treated at some point in the study time period. Leveraging a 2018 law in the state of Tennessee legalizing SSPs, a difference-in-differences analysis is utilized to compare fatal and nonfatal overdoses before and after the adoption of legalized SSPs. Due to the low number of counties adopting legalized SSPs in the state, randomization inference was used to reduce bias in the results.
Results
Both gross number of fatal and nonfatal overdoses as well as rates of overdose per 100,000 were analyzed. Gross fatal overdoses increased by roughly 37 overdoses and gross nonfatal overdoses increased by approximately 175. However, while both positive, neither fatal nor nonfatal overdose rates per 100,000 were statistically significant. 
Conclusions
Using random inference, I identify a positive and significant effect on gross fatal and nonfatal overdoses. Counties with an SSP had about 37 more fatal and 175 more nonfatal overdoses. These results are similar to a recent study looking at overdose mortality and SSPs throughout the United States. However, that paper showed a significant result in overdose death rates, and this paper only shows a significant change in the gross number of overdoses. The findings suggest some effect on overdoses at the county level when an SSP is operational.