Opioid analgesics prescribing trends 2010–2019 in Slovenia: National database study

Significant increases in global opioid use have been reported in recent decades. This study analyzed opioid utilization in outpatient care in Slovenia between 2010 and 2019.

. However, because they carry substantial risks of addiction, abuse, and other side effects, the usage of OA is strictly regulated (Berterame et al., 2016;Vranken et al., 2018).
OA use varies greatly between countries.In the United States an 'opioids crisis' due to overuse is ongoing and well documented (Centres for Disease Control and Prevention, 2020.America's drug overdose epidemic: data to action, 2020).Although in Europe the overuse cannot be compared to the crisis in the United States, use has also been increasing for the last decade in other developed countries (Bosetti et al., 2019;Karanges et al., 2018).At the same time, however, the strict regulation and increasing fear of overuse may result in a hesitancy to prescribe OA (Alqueres et al., 2015).In Slovenia, the prescription policy for OA is rather strict.There are specific "prescribing rules" for strong OA that include: duplicate of a special handwritten prescription, compulsory identification of the patient in the doctor's office and in the pharmacy, mandatory age over 18 years to allow filling the prescription, amount of OA prescribed for a maximum of 30 days per prescription, and prohibition of automatic re-prescription (Medicinal Products Act, 2014).These strict rules may affect the prevalence of OA use and their prescribing patterns.
Data on OA use in Slovenia is scarce.The International Narcotics Control Board (INCB) reported trends in the use of opioids in 25 selected European countries, including Slovenia (Häuser, Buchser, et al., 2021).The report states that the total consumption of opioids in Slovenia doubled between 2004 and 2006 and 2014 −2016.The Board also provides more detailed trends for 22 countries, excluding Slovenia; thus, little is known regarding the specifics of OA consumption and utilization trends from 2016 onwards.Moreover, few new forms and combinations of OA have entered the Slovenian market since 2010.The main novelty on the market was the combination of oxycodone, a widely used strong OA, with naloxone, an opioid antagonist.Tapentadol, a novel strong OA, was also introduced in 2012.We hypothesized that these new drugs/combinations had an impact on trends in OA use in Slovenia.
The aim of our study was to assess OA use in Slovenia between 2010 and 2019 in terms of the number of dispensed prescriptions, the number of dispensed DDDs, the number of recipients and the type of pharmaceutical formulation dispensed.

| Study design and population
This retrospective cross-sectional study analyzed national health insurance claims data containing all outpatient OA prescriptions issued in Slovenia in the period from January 1, 2010 to December 31, 2019 (the 'study period').The inclusion criterion for the study was that study participants had at least one OA prescription dispensed during the study period.OA were defined as medications with Anatomical Therapeutic Chemical (ATC) classification N02 A with the indication for treatment of pain in the SmPC.As intravenous medications are not prescribed through outpatient prescriptions, some medications from ATC N02 A were automatically excluded (piritramid, petidine).Codeine was not included in the analysis because the product marketed in Slovenia has only an indication for the treatment of cough and is not categorized under ATC N02 A (Kodeinijev fosfat alkaloid-int 30 mg tablets.Summary of Product Characteristics, 2021).This resulted in the following drugs being included: buprenorphine, dihydrocodeine, fentanyl, hydromorphone, morphine, oxycodone (as a single agent and combined with naloxone), tapentadol, tramadol (as a single agent and combined with paracetamol) (KODEINIJEV FOSFAT ALKALOID-INT 30mg tablets SmPC, 2021).Dihydrocodeine was later excluded from the analysis because only 3 prescriptions were dispensed in the whole study period.The included OA were categorized according to their strengths to weak OA (tramadol) and strong OA (buprenorphine, fentanyl, hydromorphone, morphine, oxycodone, tapentadol).

| Data sources
Health insurance claims data were obtained from the national database of the Slovenian Health Insurance Institute, the authority that manages insurance funds on behalf of the national health system in Slovenia.Slovenia has a publicly funded healthcare system that provides universal access to medical care and medications for its residents (approximately 2.1 million).The health insurance claims database lists all publicly funded outpatient prescriptions dispensed in Slovenia.None of the analyzed medications is available as an overthe-counter product.The database does not include information on sales of over-the-counter medications, hospital-dispensed medications, or privately (out-of-pocket) prescribed medications.In Slovenia, privately (out-of-pocket) prescribed medications account for less than 5% of all outpatient prescriptions (Slovenian National Health Insurance annual report 2021, 2022).Moreover, prescribing OA through private prescriptions is forbidden by law and strictly regulated (Medicinal Products Act, Official Gazzette of RS, 17/2014) More details on the Slovenian healthcare system and data on health insurance claims can be found elsewhere (Cebron Lipovec et al., 2020).
All study data were anonymized and unique patient identifiers were allocated, allowing patient-level analysis.Patient-specific study variables were: sex and age, and prescription-specific study variables were: dispensed medications, coded by the ATC classification, and date of prescription; neither the prescribed dose nor the prescribing indication was included as a study variable.

| Ethical considerations
Our study was a retrospective analysis of routinely collected data from the Slovenian Health Insurance Institute.

| Data analysis
The prevalence of OA recipients in specific age groups in the study was defined as the number of study participants ("recipients") divided by the total Slovenian population in each group in that calendar year, expressed as recipients per 1000 inhabitants.Age-standardized prevalence was calculated using the age distribution of the Slovenian population in 2010.OA consumption was presented as the number of dispensed prescriptions per 1000 inhabitants (prescriptions/1000 inhabitants) and the total number of OA defined daily doses (DDD) per 1000 inhabitants (DDD/1000 inhabitants), where DDD is defined as »the assumed average maintenance dose per day for a drug used for its main indication in adults« (WHO, 2020).For the combination of tramadol and paracetamol the DDD is defined as the number of units per day (4 units for the strength 37.5/325 mg, 2 units for the strength 75/650 mg).The DDD was therefore recalculated into mg and a DDD of 150 mg was used for both strengths of this combination.DDD per prescription was calculated by dividing the aggregated number of dispensed DDD by the aggregated number of dispensed prescriptions in each year.This was defined as the "prescribing intensity".
The proportion of dispensed pharmaceutical formulations was reported as the percentage of DDD for 3 categories: modifiedrelease oral form, immediate-release oral form and transdermal patches (TDP).Modified-release oral forms included prolonged-

| Prevalence of OA recipients
Our study included an average of 145,500 opioid recipients analyzed each year.The average age was between 63 years (2010) and 66 years ( 2019) and the number of female recipients was about 1.5 the number of male recipients throughout the study period.Recipients of tramadol represented between 91% and 93% of all recipients in all years analyzed.
The overall prevalence of OA recipients increased from 7.0% in 2010 to 7.8% in 2014 and then steadily dropped to 6.5% in 2019 (not shown in Figure 1).The age-standardized prevalence was 5.5% in 2019, resulting in a relative prevalence change of 21.5% (Figure 1a).
The trend in the prevalence of tramadol recipients closely mimicked the trend of all recipients, whereas the prevalence of strong OA recipients was much lower (between 0.5% and 0.7%, Figure 1b).
Prevalence was highest among older patients (>65 years), where OA were prescribed to more than 20% of patients between 2010 and 2017 and decreased to 17.5% in 2019.The prevalence in this age group was substantially higher compared to adults (3 −4x for tramadol, 10x for strong OA).(Figures 1a and 1b).inhabitants followed a similar trend, but the overall decrease was larger (9.2%).The consumption of tramadol followed a similar trend (2.7% decrease in DDD, 12.2% decrease in prescriptions).The prescribing intensity (number of DDD per prescription) of tramadol increased throughout the study period.This is mainly attributed to the intensification of consumption of tramadol/paracetamol (þ57% DDD, þ10% prescriptions).However, single agent tramadol consumption decreased by more than 50% both in terms of the number of prescriptions and the number of DDDs (Table 1 and Table 2).

| Consumption of OA in DDD and prescriptions per 1000 inhabitants
The consumption of strong OA in DDD represented between 18% and 21% of all OA consumption.The consumption slightly increased between 2010 and 2014 (6.7%) and decreased by more than a fifth between 2014 and 2019 (21,3%; overall decrease 16,2%).
In contrast to tramadol, the number of prescriptions for strong opioids increased by 10.2%.The change in DDD/prescription shows that the prescribing intensity of strong OA decreased during the observed period (Tables 1-3).
Of the strong opioids, the most frequently prescribed were fentanyl and oxycodone, representing more than 70% of all strong OA both in terms of dispensed prescriptions and DDD throughout the study period.However, their consumption changed significantly during the observed years.The consumption of fentanyl exceeded 280 DDD/1000 inhabitants between 2010 and 2014 and dropped by 2019 by more than a third to 184 DDD/1000 inhabitants (overall decrease 38%).The number of prescriptions of fentanyl also decreased (28%), but less than the number of DDD.On the contrary, the consumption of oxycodone increased by 50% between 2010 and 2016 (from 95 DDD/1000 inhabitants to 151 DDD/1000 inhabitants) and dropped slightly between 2016 and 2019 (overall increase 35%).The increase was mainly due to the increased consumption of oxycodone with naloxone, whereas single agent oxycodone decreased for almost 50% in terms of DDD.In terms of number of prescriptions, the consumption for oxycodone more than doubled by 2016 and dropped slightly by 2019 (overall increase 74%).The change in DDD per prescription shows that the prescribing intensity of both fentanyl and oxycodone decreased during the observed period.Other strong opioids were prescribed less frequently.The highest increase both in the number of DDD and the number of prescriptions was observed for tapentadol, which appeared on the Slovenian market in 2012 for the first time.The remaining strong opioids (buprenorphine, hydromorphone and morphine) were used at rates below 60 DDD/1000 inhabitants per year and showed a decrease in DDD and a steady or decreasing number of prescriptions dispensed (Table 3).

| Consumption of different pharmaceutical forms of OA
The consumption of opioids in terms of their pharmaceutical form revealed changes both for tramadol and strong OA.In 2010, tramadol was prescribed either as modified-release or immediaterelease oral forms, in similar proportions (44.3% vs. 55.5%,respectively).By 2019 however, the consumption of immediate-release forms had increased significantly and represented more than 80% All OA 13.3 13.2 12.9 12.9 12.7 13.0 13.6 14.0 13.9 13.9 Note: The values within each category (tramadol, strong opioids) separately are represented on the color scale from lowest (green), medium (yellow) to highest (red).The gray shade is meant to show the categories of tramadol, strong AO and all OA.The gray-unshaded lines constitute these categories.
CEBRON LIPOVEC of all tramadol.Strong OA were prescribed in oral forms (modifiedrelease, immediate-release) or transdermal patches.In 2010, the most commonly prescribed forms were transdermal patches (67.6%, mostly fentanyl).This steadily decreased to 46.8% in 2019 on account of the increased use of modified-release oral forms.The use of immediate-release oral forms also increased but constituted at most one 10th of all strong OA consumption (Figure 2).

| DISCUSSION
The number of OA recipients in Slovenia is decreasing, whilst the overall consumption is either relatively stable (tramadol) or is decreasing (strong OA).Consumption is highest among older patients, where an OA is prescribed to almost one in 6 patients.Of the strong OA, oxycodone (in combination with naloxone) and fentanyl are now prevailing, but are being prescribed less intensely (less DDD per prescription) than 10 years ago.To our knowledge, this is the first study reporting a detailed 10-year trend of OA use in Slovenia.
Comparison of our results with other studies is challenging because of differences in methodology, lack of comparable numeric data and major differences in the duration, setting, and timeline of studies, so comparisons and conclusions should be interpreted with caution.Previous reports showed differences in prevalence and consumption of AO among European regions and countries.The reported annual prevalence is lower than the reported prevalence in Ireland (Norris et al., 2021), and Finland (Muller et al., 2019), but higher than the reported prevalence in Estonia (Uusküla et al., 2020).
In terms of consumption, our results report a DID (DDD per 1000 inhabitants per day) between 7.1 and 6.7 for all OA, for tramadol between 5.7 and 5.6 and for strong opioids between 1.5 and 1.2 DID, which is 10x lower compared to Ireland (Norris et al., 2021) and 5x lower compared to Spain (Hurtado et al., 2020).However, our results Europe, but substantially higher than the consumption in Eastern European countries (Ju et al., 2022).
We observed interesting trends within the observed study period: a general increase between 2010 and 2016, followed by a decline up to 2019.Previous studies assessing OA use in various European countries reported an increase in OA use in Slovenia up to 2016 (Bosetti et al., 2019), which is in line with our data.Similarly, a recent study by Ju et al. reporting global, regional, and national trends in OA consumption between 2015 and 2019 showed a 5% decrease in OA consumption per year in Slovenia (evaluated in MME (Miligram Morphine Equivalent) per 1000 inhabitants per day) (Ju et al., 2022).Interestingly, in 2016, the UN General Assembly on the World Drug Problem called on all governments to improve the control and supply of OA to combat the global problem of overuse or underuse of OA (UNODC, 2016).
Our findings also shed light on some interesting disparities on the use of tramadol and strong OA.The use of tramadol far exceeded the use of strong OA in all observed measures.This mirrors the upward trends of tramadol use in several other countries, such as Scotland (Ruscitto et al., 2015), Norway (Muller et al., 2019), Spain (Hurtado et al., 2020), Ireland (Norris et al., 2021), and Taiwan (Chen et al., 2021).The results are also in line with previously published national reports of OA use (Bedene et al., 2021).In Slovenia however, the use of single agent tramadol is decreasing and is being replaced by tramadol in combination with paracetamol, which was also observed by Norris et al. (2021) The combination is being prescribed more intensively than 10 years ago, implying longer treatment duration or higher prescribed dosages.
The increase could also reflect the hesitance to prescribe single agent tramadol.In fact, concerns about the abuse and misuse of tramadol have emerged in recent years as there is increasing evidence that it is being misused as a drug and has been implicated in For strong opioids we have seen a decline in the prevalence of recipients as well as consumption in DDD, but an increase in the number of prescriptions.From our data, it seems that strong opioids are being prescribed more cautiously (less DDD per prescription) than 10 years ago and to fewer patients.The decrease in prevalence and consumption is also in contrast to results from numerous European countries (Bosetti et al., 2019;Engi et al., 2022;Hurtado et al., 2020;Norris et al., 2021;Rosner et al., 2019)  Fentanyl was and remains the most widely prescribed strong OA in an outpatient setting in Slovenia as well as in numerous European countries (Engi et al., 2022;Hurtado et al., 2020;Norris et al., 2021;Rosner et al., 2019).However, these studies do not report a decrease in fentanyl use as was observed in our study.As fentanyl is mostly prescribed in the form of TDP, this also resulted in a decreased proportion of use of this pharmaceutical form.This result is in contrast to observations from the abovementioned studies, where the use of transdermal forms is increasing.Although TDPs have numerous advantages over oral formulations (use in dysphagia, decreased constipation, patient-friendly use), their inappropriate use can also pose significant risks and result in adverse events and even death (Cheema et al., 2020).Their use was replaced by modifiedrelease oral forms, mostly oxycodone with naloxone, which, due to the antagonistic function of naloxone, acts as a safer choice and is classified as an abuse-deterrent OA (Morlion et al., 2018).Our finding of increased oxycodone prescribing is consistent with a large number of other EU studies which report a marked increase in oxycodone consumption over the past 2 decades (Curtis et al., 2019;Hider-Mlynarz et al., 2018;Muller et al., 2019;Norris et al., 2021).
However, many of those studies do not differentiate between single agent oxycodone and oxycodone in combination with naloxone.A substantial increase in tapentadol consumption was also observed, also in line with reports from other European countries (Curtis et al., 2019;Engi et al., 2022;Hurtado et al., 2020;Norris et al., 2021;Rosner et al., 2019) (Häuser, Morlion, et al., 2021).Similar to guidelines in the United States, European guidelines have moved away from recommending opioids as first-line treatments for chronic non-cancer pain (Dowell et al., 2016;Häuser, Morlion, et al., 2021).This could have resulted in generally more cautious prescribing of OA, especially for non-cancer pain.
Guidelines also recommend the use of prolonged-release formulations only when prolonged treatment is necessary and emphasize their potential for side-effects (CDC, 2018  , 2016).In 2018, the modified-release tramadol with paracetamol 75 mg/650 mg was temporarily withdrawn from the Slovenian market (Krka d.d., 2018).The latest could have increased the consumption of lower strength tramadol with paracetamol (increasing the number of prescriptions and DDD/prescription).Last but not least, the arrival of generic brands to the market (eg.generic fentanyl TDP in 2017, generic oxycodone in 2013) could have changed the marketing strategy for the innovative drugs and impacted the overall prescribing rate.

| Strengths and limitations
Our study is highly comprehensive as it used a large database that provided nationwide data covering the complete outpatient use of OA.Moreover, our study provides a detailed analysis of opioid use over a 10-year period, including the differentiation between single agent and combinational tramadol and oxycodone as well as the analysis of consumption of different pharmaceutical forms.Nonetheless, some limitations need to be considered.Firstly, our data sets did not contain information on the prescribed dose, duration of treatment and therapeutic indication.This made it impossible to CEBRON LIPOVEC -7 of 9 differentiate between chronic cancer pain, chronic non-cancer pain and acute pain management and hence to appraise the quality of prescribing.Further, our results report ambulatory OA use.As OA are often used in palliative care and in the hospital setting, our results cannot be compared to hospital OA use or overall OA use (eg, sales analysis results).Generalization of our findings should therefore be approached with caution.

| CONCLUSIONS
The release tablets and prolonged-release capsules.Immediate-release oral forms included film-coated tablets, hard capsules, oral solution, oral drops, sublingual tablets and effervescent tablets.This analysis was performed for the years 2010, 2015 and 2019.Study participants were placed into the following age groups: <18 years, 18−65 years (adults) and >65 years (older adults).All statistical analyses were conducted using IBM SPSS Statistics 27.0.
Overall OA consumption in terms of prescribed DDD increased by 6.3% between 2010 and 2016 (from 2599 DDD/1000 inhabitants to 2763 DDD/1000 inhabitants), and then dropped by 11.0% by 2019 (2458 DDD/1000 inhabitants), resulting in an overall decrease of 5.4% during the study period.The number of prescriptions per 1000 F I G U R E 1 Prevalence of OA recipients between 2010 and 2019.(a) All OA recipients.(b) Strong OA recipients.Note that in (b) the scale is 10 times smaller than in (a).The age group below 18 years is not presented in the figure because the prevalence was extremely low during the whole study period (1.3 recipient/1000 inhabitants for weak OA, 0.04 recipient/1000 inhabitants for strong OA).
report a higher consumption compared to Estonia, and seem to be close to the recently published data from Hungary (Engi et al., 2022; Uusküla et al., 2020) A similar regional division was observed by Ju et al., where the overall consumption of OA (ambulatory and hospital use) in Slovenia is half of the consumption in Western and Central overdose deaths (WHO, 2018).The WHO Expert Committee on Drug Dependence has acknowledged these concerns and stated that tramadol could be classified as a narcotic under the Single Convention on Narcotic Drugs (WHO, 2018).F I G U R E 2 Proportion of different pharmaceutical forms dispensed in 2010, 2015 and 2019.(a) Tramadol.(b) Strong OA.TDP: transdermal patches.Tramadol was also dispensed in rectal form (2010: 0.2%, 2015: 0.0%, 2019: 0.2%).Strong OA were also dispensed in nasal form (2015 only: <0.1%) and buccal form (2010: 0.4%, 2015: 1.4%, 2019: 1.6%).
and might imply a more cautious use of strong OA.On the other hand, the low prevalence of strong OA recipients (3.6% in 2019), especially among the oldest patients (2% in 2019) combined with the aging of the population and an increasing number of cancer patients and patients in palliative care might also imply OA underuse and raises the question whether the basic needs for pain control are being met in Slovene patients.Further research is urgently needed to understand the implications of decreased use of strong OA in Slovenia.
T A B L E 1Note: The values within each category (tramadol, strong opioids) separately are represented on the color scale from lowest (green), medium (yellow) to highest (red).The gray shade is meant to show the categories of tramadol, strong AO and all OA.The gray-unshaded lines constitute these categories.aTheoverall change for tapentadol and oxycodone with naloxone is calculated from 2013 (first full year on the market) to 2019.Note:The values within each category (tramadol, strong opioids) separately are represented on the color scale from lowest (green), medium (yellow) to highest (red).The gray shade is meant to show the categories of tramadol, strong AO and all OA.The gray-unshaded lines constitute these categories.a The overall change for tapentadol and oxycodone with naloxone is calculated from 2013 (first full year on the market) to 2019.

19.6 19.1 17.8 17.0 16.4 15.9 15.6 15.2 15.1
These emphasize a multimodal approach to pain management.This approach combines various non-opioid treatments, such as physical therapy, psychological interventions, and non-opioid medications, to reduce reliance on opioids substantial changes in the national guidelines.In fact, the latest guidelines for non-cancer pain treatment date in 2007 (Krčevski-Škvarč N, 2007)and for cancer treatment date to 2015(Lahajnar   Čavlovič S et al., 2015)with no major changes compared to the previous versions.However, changes have been observed in international guidelines over the past decade, especially in guidelines related to non-cancer pain.
prevalence of OA recipients in Slovenia decreased between 2010 and 2019 (7% decrease in absolute prevalence, 21.5% decrease in age-standardized prevalence)-from 7.0% to 5.5% -and is considerably lower compared to the European average.The overall consumption of OA also decreased; however, a closer look revealed that tramadol consumption was relatively stable, whereas the consumption of strong OA decreased significantly.Of the strong OA, oxycodone (in combination with naloxone) and fentanyl prevailed throughout the observed period, but are being prescribed less intensely than in 2010.Overall OA consumption in Slovenia is also substantially lower than the reported European average.Further research is warranted to understand whether the low and further decreasing use of OA represents safe use or underuse of these important analgesic drugs.