In Reply As Dr McAuliffe notes, estimates of the risk of problematic opioid use associated with opioid prescription have varied widely depending on a number of factors including outcome measured, methods used to ascertain the outcome, and population studied. In our article, we reported estimates of the prevalence of DSM-IV opioid dependence from 3 studies of primary care patients on long-term opioid therapy in which opioid dependence was assessed using diagnostic interviews. As McAuliffe suggests, estimates for opioid dependence might underestimate rates of opioid use disorder. Opioid use disorder as a single, unifying diagnosis with different levels of severity is a relatively recent construct introduced in DSM-5 to replace the 2 DSM-IV diagnoses of opioid dependence and opioid abuse. Few studies to date have assessed opioid use disorder using DSM-5 criteria. However, estimates for DSM-IV opioid dependence and opioid abuse cannot simply be combined into an estimate for DSM-5 opioid use disorder, as McAuliffe recommends, because the diagnostic criteria do not align completely.1 Although the studies’ different time frames for meeting criteria for opioid dependence likely affected the studies’ results, this was less relevant given that we were reporting prevalence and not incidence, and we reported on studies in which participants had been receiving opioids for at least a year or close to a year. Ideally future studies will assess opioid use disorder using DSM-5 criteria and systematic diagnostic interviews.
Dowell D, Haegerich TM, Chou R. Prescribing Opioids for Chronic Pain—Reply. JAMA. 2016;316(7):774-775. doi:10.1001/jama.2016.9242