In Reply The data presented by Gambatese and colleagues from New York City, New York, and Riley and colleagues from San Francisco, California, suggest that the epidemic of drug overdose deaths among homeless individuals is not confined to Boston, Massachusetts. We are encouraged by the concordance of our results with their findings but saddened by the scale and scope of this problem.
We agree with Riley and colleagues that this issue requires a coordinated and comprehensive public health response. Since the publication of our study, Boston Health Care for the Homeless Program has committed to doubling the size of its buprenorphine treatment program, expanded the distribution of nasal naloxone to persons at risk of opioid overdose, and joined with public health officials and community stakeholders to participate in a citywide overdose prevention task force. While opioids were the chief contributors to overdose deaths in our study and in the data reported by Gambatese et al, Riley and colleagues report that cocaine was the principal substance implicated in overdose deaths in their cohort of human immunodeficiency virus–infected women. Although less common in our cohort, cocaine contributed to more than one-third of overdose deaths, and multiple drugs were implicated in 43% of overdoses. These findings support the call by Riley et al for overdose prevention programs that address polysubstance use.