Opioid and analgesic utilization in Ireland in 2000 and 2015: A repeated cross‐sectional study

Abstract In recent decades, opioid use has increased internationally and is a major public health concern. This study aims to characterize changes in opioid and other analgesic prescribing in Ireland over a 15‐year period (2000–2015). This is a repeated cross‐sectional study of administrative pharmacy claims data in 2000 and 2015. Individuals of all ages in Ireland's Eastern Health Board region who were eligible for the General Medical Services (GMS) scheme were included. This scheme covers 40% of the population, mostly those on lower incomes and older people. The primary outcome was dispensing of opioids, both prevalence of any use and rate per 1000 GMS eligible population (standardized to the 2015 population). Logistic regression was used to assess odds of opioid dispensing in 2015 versus 2000, controlling for demographic differences. The eligible study population was 364 436 in 2000 and 523 653 in 2015. In 2000, 19.4% of the eligible population had at least one opioid dispensing compared to 20.8% in 2015. The rate increased from 671 to 1098 dispensings per 1000 population. The increase was highest in the dispensing rates of codeine, tramadol, oxycodone, buprenorphine, and fentanyl. Compared to 2000, there was higher odds in 2015 of being dispensed a strong opioid (adjusted odds ratio 2.0, 95%CI 1.97–2.04) or long‐acting formulation (3.75, 95%CI 3.58–3.92). Increased prescribing of opioids, particularly strong opioids, between 2000 and 2015 is evident in Ireland. This is concerning due to the potential for misuse, and opioid‐related morbidity/mortality.

extent that in the United States this has been termed the opioid epidemic. 1 Although opioids are a mainstay in treating acute and cancerrelated pain, they are increasingly being used, and now predominantly so, for chronic non-cancer pain where benefits are less likely to outweigh risks. 2 A recent US study found opioid-attributable deaths has increased threefold from 2001 to 2016, equivalent to 1.68 million person-years of life lost in 2016. 3 In England, prescribing of morphine-equivalent opioids increased by 127% between 1998 and 2016. 4 In Ireland, although codeine prescribing has been examined, 5 there has been little evidence on trends in opioid prescribing.
However, two recent studies have suggested Ireland has relatively high opioid overdose mortality, and that prescribing of strong opioids has increased between 2010 and 2019. 6,7 With changing opioids prescribing, it is important to understand if changes are driven by higher prescribing rates or doses/potencies, and also how this fits within the context of other analgesic medications. The potential for inappropriate use and misuse of analgesics is not limited to opioids, with treatments for neuropathic pain, such as pregabalin and gabapentin, also of concern. 8,9 This has resulted in the reclassification of pregabalin as a controlled drug in the UK in 2019 to reduce potential misuse and abuse. 9 Lastly, concerns have been raised that prescribing of lidocaine patches, a recently developed analgesic product approved for post-herpetic neuralgia, outside of their approved indication represents low-value care based on cost and lack of robust evidence of effectiveness. 10 Therefore, this study aims to characterize changes in analgesic, and specifically opioid, prescribing in Ireland over a 15-year period from 2000 to 2015.

| MATERIAL S AND ME THODS
This is a repeated cross-sectional study based on secondary analysis of administrative dispensing data. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement was used to guide the reporting of this manuscript. 11 The setting is primary care in the Eastern Health Board (EHB) region of Ireland, which is the largest of eight regions including 29.3% of the national population. The study focuses on 2 years, 2000 and 2015, being the earliest and latest full years for which data were available when data analysis commenced in 2017, and we did not have access to data for intervening years. Data used in this study were provided by the Health Service Executive Primary Care Reimbursement Service (HSE-PCRS) at this time. No formal mechanism of access to item-level data exists currently. Ireland has a mixed public private health system, with approximately 40% of the population having entitlement to a range of public health services, including medicines, at low or no cost through the General Medical Services (GMS) scheme.  (Table S1). This did not include methadone or specific buprenorphine formulations dis- Opioid-paracetamol combinations were classified and analyzed based on their opioid ingredient. Although other medications can be used in the treatment of pain, the research team judged that the above drug could be reliably assumed to be prescribed for analgesia, rather than alternative indications.

This study includes individuals of all ages in the EHB region of
Opioid medications were further examined in terms of strong opioids (as per the British National Formulary Classification, see Table S1), long-acting formulations (patches or modified-release tablets), and individual opioid drugs. These were summarized in terms of rate of dispensings and Oral Morphine Equivalents (OME) per 1000 population, also standardized to the 2015 GMS population. OMEs

What is already known on this subject?
• Opioid-related morbidity and mortality are a significant public health concern internationally, partly driven by the prescription of opioid medications.
• Other analgesic medications, such as gabapentinoids, have also been implicated in drug-related deaths.
• To date, there is limited evidence on trends in prescription of opioids and other analgesics in Ireland.

What this study adds?
• Between 2000 and 2015, the prescription of opioids has increased in Ireland, with particularly sharp increases in the prescription of strong opioids and long-acting formulations.
• Wider availability of prescribed opioids to address pain needs to be balanced against known medication-related harms.
for each dispensing were calculated by the product of the quantity, strength, and conversion factor. 4 Logistic regression was used to determine the odds of an opioid dispensing in 2015 compared to 2000, controlling for age group and sex, as well as the odds of a strong or long-acting formulation opioid dispensing, yielding adjusted odds ratios (aOR) with 95% confidence intervals (CI). Analysis was at the individual level, including an observation for each GMS-eligible individual. The outcome variables were binary indicators of any dispensing of an opioid, strong opioid, and long-acting formulation. Among those with an opioid dispensing, multivariate negative binomial regression was used to assess change in the rate of dispensings and OMEs between 2000 and 2015, yielding adjusted rate ratios (aRR) and 95% CIs. Statistical analysis was conducted using Stata version 14, and statistical significance was assumed at p < .05, and analytical code is available at www.doi. org/10.5281/zenodo.5570200.  (Table S3) Analgesic (any of the below classes/agents) greater proportional growth in some higher potency opioid agents, for example, oxycodone, fentanyl, and buprenorphine. Although the largest increase in the rate of dispensings was for codeine, it accounted for less of an increase in OMEs given its low potency. Decreasing use was observed for dextropropoxyphene, the opioid ingredient in coproxamol which was withdrawn from the Irish market in 2006, as well as morphine and dihydrocodeine. The decline in morphine concurrent with growth in newer opioids has been reported in other studies, as prescribers switch to these as the strong opioid of choice. 4,6,12 Promotion and marketing of newer opioids and opioids formulations may have contributed to uptake of these agents. per 1000 population in the United States, 17 although with noted variation between provinces/states. While our findings are higher than many of these, this may reflect that the GMS scheme overrepresents socioeconomically deprived and older individuals, among who painful conditions may be more prevalent. 18

| Strengths and limitations
This study was limited to dispensing of medicines prescribed to GMS in particular opioid, use in a general primary care population. Although more recent data than 2015 on strong opioids have been analyzed elsewhere, 6 our study provides an earlier historical benchmark to understand analgesic prescribing changes. It also characterizes opioid utilization using morphine equivalents, a robust measure to evaluate shifts in the quantity and potency of opioid prescribing. 19

| CON CLUS IONS
Increasing prescription use of opioids, particularly strong opioids and long-acting formulations, and gabapentinoids is concerning due to the potential for misuse, diversion and opioid-related death. Such concerns are reflected in reports of increasing drug-related deaths in Ireland. 20 However, restrictions to reduce drug-related harm must be balanced against the need to ensure access to appropriate pain relief. Future research should examine the trajectory of opioid dispensing trends in Ireland, and how this relates to medication-related morbidity and mortality.

ACK N OWLED G EM ENTS
We wish to acknowledge the Health Service Executive Primary Care Reimbursement Service for access to the anonymous data used in this study.

D I SCLOS U R E
None declared.

E TH I C S A PPROVA L
This study used anonymized data from pharmacy claims and therefore did not require ethical approval.