Prevalence and Incidence Trends for Diagnosed Prescription Opioid Use Disorders in the United Kingdom

The increase in the prescribing of opioid analgesics for the management of pain has been paralleled by increases in adverse outcomes associated with misuse and abuse, particularly in the United States [1]. Despite increased prescribing of opioid analgesics in Europe [2], few studies have sought to investigate the extent of this relationship [3]. In a recent review of prescription opioid use disorders by Vowles et al. [4], of the 38 studies included, only two were from Europe. Because the US is not comparable to most European countries in terms of the health care system and prescription practices [5], the strength of the relationship between prescription opioid use and subsequent risk of misuse and abuse seen in the US should not be assumed to be generalizable beyond North America.

Similar to the US, the UK has seen a marked increase in opioid prescribing in the past decade, especially of the stronger opioids [6]. In contrast to the US, however, there is limited information on the relationship between prescription opioid use and the development of opioid use disorders in the UK [4, 5]. As such, using data from the UK Clinical Practice Research Datalink (CPRD), the largest database of anonymized, longitudinal primary care records including approximately 4.4 million active patients, we aimed to describe the characteristics of patients with a diagnosis of prescription opioid use disorders and investigate the 5-year cumulative prevalence and trends in incidence rates of prescription opioid use disorders between 2008 and 2012. Additionally, we sought to examine the prevalence and incidence of opioid use disorders stratified by the different opioids available in the UK.

In this large population-based cohort study of prescription opioid users followed for up to 5 years, we found that the prevalence of opioid use disorders was low at 0.05% and that there was no obvious change in the incidence of newly diagnosed opioid use disorders over the duration of the study. These findings challenge the long-standing belief that increases in prescribing of opioids necessitates an increase in the incidence of problematic opioid use disorders. Our findings also demonstrate that it is the most commonly first-prescribed opioids, and not necessarily the more potent opioids, which are most frequently associated with an opioid use disorder in the 6 months leading up to diagnosis of an opioid use disorder.

As far as we are aware, this is the first nationally representative medical records database study which has sought to investigate the association between opioid use and risk of subsequent opioid use disorders in the UK. A recent systematic review of rates of opioid misuse, abuse, and addiction in chronic pain included a total of 38 studies conducted mostly in the primary care setting or pain clinics [4]. Across most of the calculations, rates of misuse averaged between 21% and 29% (range 95% CI 13%–38%). Of note, the majority of the included studies were not database studies and the definitions used for misuse were not consistent across studies. Furthermore, out of the 38 included studies, only two were from Europe (UK and Denmark), and neither of these studies included rates for misuse or abuse; only one of the 38 studies included an estimate of opioid abuse. A study using UK death data from the Office for National Statistics has shown an increasing trend in the number of deaths attributable to opioids from 2001 to 2011 [15]. However, the results in this study were expressed as absolute number of deaths and did not account for increases in opioid prescribing over the same duration. Furthermore, using mortality data as a proxy for rates of misuse/abuse is problematic given that many other factors other than misuse/abuse are known to influence risk of opioid overdose and death, e.g., opioid dose, co-prescription with benzodiazepines, time to administration of an opioid antagonist (i.e., naloxone) and response time by emergency services. A recent study has shown that the prevalence of past-year misuse/abuse of prescription opioids in Great Britain was 6.2% [16]. The high prevalence of opioid misuse/abuse found in that study should however be interpreted with caution since the participants were not drawn from a representative nationwide database but were purposely pre-selected for problematic use, i.e., they were aged 12–49 years, likely to use tobacco and marijuana, and were recruited from needle exchanges, homeless shelters, and parks.

In analyses stratified by opioid drug, we found that the relation between opioid use disorders and prescriptions of codeine and tramadol remained stable, whereas there was an increasing trend for morphine and a decreasing trend related to dihydrocodeine. In a previous study examining a cohort of new opioid users who started treatment with weak opioids (i.e., codeine, tramadol, and dextropropoxyphene), Skurtveit et al. [17] found that 0.08% of patients went on to develop prescription patterns indicative of an opioid use disorder. This is consistent with our observation for morphine which, despite being highlighted as being associated with an increasing incidence during the study, was 0.05%. Skurtveit et al. did not examine associations with more potent opioids or morphine, which limits direct comparison with our findings [17]. We were not able to examine for trends in opioid use disorders for buprenorphine, oxycodone, fentanyl, metazinol, diamorphine, dipipanone, hydromorphone, methadone, papaveretum, pentazocine, pethidine, and tapentadol due to too few cases. This is important because whilst our findings suggest that there is unlikely to be an abuse problem with these specific opioids at this time, it is not to say that this will not change in the future. This highlights the importance for ongoing surveillance of this important public health issue.

Important strengths of our study include use of a large representative population database with up to 5 years of follow-up data and diagnosis of opioid use disorders by family physicians. Additionally, use of a publicly available medical records database, inclusion of an in-depth statistical analysis strategy, and inclusion of a participant flow diagram, means that our findings can be replicated and repeated in future analyses to determine trends in opioid use disorders over extended durations of time.

There are a number of limitations of our study which warrant discussion. First, whilst a number of validation studies have demonstrated the high validity of the diagnoses codes recorded in CPRD, reporting strong measures of sensitivity and specificity [7, 10‐12], this has not been undertaken for the codes related to problematic opioid use disorders. Secondly, data on medications given during hospitalization, medications provided in specialist care, and medications provided by a hospital following patient discharge are not recorded in patients’ medical records. The likely impact of this will have been an underestimation of drug use and, potentially, an underestimation of the extent of opioid use disorders. Nevertheless, under the assumption that these factors did not change during the study period, an examination of annual incidence rates does allow an accurate estimation of changes in trends over time even if the absolute number of patients is not estimated with precision. Thirdly, although we excluded patients who received an opioid prescription used for the purpose of substitution therapy, some of the opioid use disorder cases in our study may have received a diagnosis of an opioid use disorder based on illicit opioid use. This would, however, have led rather to an overestimation of prevalence and incidence. Finally, the possibility of significant under-diagnosis by patient’s physicians does mean that future research using alternative data sources is needed, such as the research being conducted by the RADARs group in the US who recently examined prescription opioid misuse and abuse using multiple data sources in a complimentary analysis approach e.g., drug-diversion investigations, poison centers and substance-abuse treatment centers [18].

In conclusion, despite the marked increase in overall opioid prescribing in the UK in the past decade, especially of the stronger opioids, this has not lead to an increase in the incidence of diagnosed opioid use disorders. Given the potential for under-reporting of prescription opioid use disorders in CPRD, and the likely important influence that differing prescribing patterns and health care systems might play in determining the development of opioid use disorders, future longitudinal studies using different data sources across Europe are needed to further our understanding of this important public health issue.