INTRODUCTION

The increasing availability of opioid analgesics has been accompanied by an alarming rise in opioid misuse and related harm. Worldwide, opioid utilisation increased 2.5-fold between 2001-03 and 2011-13,[1] with the United States and Canada having the largest per capita opioid consumption. Opioid-related deaths rose more than 20% in developed countries in the same period. In 2018, 46,800 people died due to opioid overdoses in the United States with 32% involving prescription opioids. [2, 3]
Opioid analgesics were initially introduced to manage acute and cancer-related pain and subsequently registered to treat chronic noncancer pain (CNCP). [4] Short-term opioid use for CNCP has limited to moderate clinical benefit [5, 6] and there is limited evidence of long-term effectiveness.[5, 7, 8] Prolonged opioid use has been associated with adverse events, hospitalisations, dependence and overdose.[9-12] Therefore, guidelines for CNCP encourage initiating therapy with low-doses and limiting treatment to no more than eight to 12 weeks. [9, 13]
Routinely collected data and drug monitoring programs, using dispensing claims and prescribing data, represent a rich source of information supporting large-scale population-based research to quantify, monitor and develop evidence-informed strategies to reduce long-term opioid therapy (LTOT). However, translating guideline recommendations to measurement in routinely collected data can be challenging. Despite the growing literature of opioid use and harms, there is no gold standard definition of LTOT for use in routinely collected data.
Recent systematic reviews have identified between 29 and 41 variants of definitions. [14, 15] These studies did not focus exclusively on definition applied to routinely collected dispensing and prescribing data. Rather they included studies ascertaining opioid use across a broad spectrum of methods including surveys. They also focused mostly on the use of opioids in surgical settings and did not shed light on the rationale for the various definitions operationalised in the studies.
Understanding the terminologies and definitions currently in place in observational research is crucial for planning research, evaluating findings, and informing evidence-based recommendations. Given the increasing role of routinely collected data to evaluate opioid use and outcomes, especially for CNCP management, we aimed to catalogue the definitions and terminology relating to LTOT, the rationale for definitions, and LTOT rates in observational research using routinely collected data in non-surgical settings.