Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prescription opioid dispensing and prescription opioid poisoning: Population data from Victoria, Australia 2006 to 2013

Prescription opioid dispensing and prescription opioid poisoning: Population data from Victoria,... been referred to as epidemic in the Objective: To describe recent trends in opioid prescribing and prescription opioid poisoning PUnited States (US). Between 2001 resulting in hospitalisation or death in Victoria, Australia. and 2014 there was a three-fold increase in Method: This is a population-based ecological study of residents of Victoria, 2006 -14. deaths from this cause, plateauing in 2011 at 2-4 Australian Bureau of Statistics residential population data were combined with Pharmaceutical about 17,000 deaths annually. US health Benefits Scheme (PBS) opioid prescription data, Victorian Admitted Episodes Data (VAED) and agencies have recognised the national cause of death data. increase in opioid prescriptions as a key driver Results: Annual opioid dispensings increased by 78% in 2006 -13, from 0.33 to 0.58 per of the increase in prescription drug overdose population. Opioid use increased with age: in 2013, 14% of Victorian residents aged ≥65 years deaths; particularly the prescribing of long- 5-9 filled at least one oxycodone prescription. In 2006 -14, prescription opioid related hospital term opioids for chronic non-cancer pain. admissions increased by 6.8% per year, from 107 to 187 /1,000,000 person-years; 56% were A number of US states have implemented due to intentional self-poisoning. Annual deaths increased from 21 to 28 /1,000,000 persons, in a series of policy responses focused on 2007 -11. Admissions and deaths peaked at 25-44 years. restriction of prescription opioid supply and Conclusions: Although both opioid prescribing and poisoning have increased, there is increased availability of the opioid antagonist 10,11 discrepancy between the exposed group (dispensings increased with age) and those with drug naloxone. While there has been adverse consequences (rates peaked at ages 25-44 years). a plateau in the non-fentanyl prescription Implications: A better understanding is needed of drivers of prescribing and adverse opioid epidemic curve that has coincided consequences. Together with monitoring of prescribing and poisoning, this will facilitate early with the constraints on prescription supply, detection and prevention of a public health problem. heroin deaths in the US since 2011 have Key words: prescription opioids, drug overdose, pharmacoepidemiology tripled, with consequent ongoing escalation of the overall opioid-related public health 2,3,12,13 problem. Fentanyl can be obtained currently have a death rate that is increasing appears somewhat resistant to intervention. on prescription or manufactured illicitly: more quickly than among men. The highest There are early indications that Australia has on the death certificate, deaths involving rate of deaths in both sexes is in the 45-55 age a prescription opioid problem. Annual opioid prescription fentanyl cannot be distinguished group. Individuals at highest risk of opioid analgesic use has increased in Australia 2002- from deaths involving illicitly manufactured related harm are those being prescribed long 2009, from 13 to 16 defined daily dosages fentanyl. The availability of illicitly term, high Morphine Equivalent Doses and (DDD) per 1,000 population per day. A study manufactured fentanyl has increased in the those obtaining concurrent prescriptions of opioid prescribing and adverse outcomes 14 17,18 US. In the US, the relationship between from different sources. There are three key in Australia reported a 15-fold increase in opioid prescriptions and opioid deaths has lessons to be learned from the US experience. opioid prescribing between 1992 and 2012, been explored in some detail. Regular opioid First, there is a substantial period between the but with an apparently small effect on opioid use in the US is more commonly reported beginning of the epidemic and the point at related harm. Oxycodone and fentanyl among women than men, with women more which a national response can be mobilised. prescribing have both increased since 2000, likely to use opioids chronically at higher Second, the increase in opioid deaths appears but unlike the US, the rate of increased doses and to increase their use with age. to be related to an increase in the amount prescribing was not shown to match the rate 21-23 Deaths are more frequent among men than of opioids prescribed, both at a population of increase in deaths. Fentanyl prescribing women, although women are hospitalised level and in relation to the group for whom in Australia has increased in recent years, for opioid overdose more frequently and prescriptions are intended. Third, the problem particularly among persons aged over 1. Monash Injury Research Institute, Monash University, Victoria 2. Institute for Safety, Compensation and Recovery Research, Monash University, Victoria 3. Harvard Injury Control Research Center, Harvard School of Population Health, Massachusetts, USA Correspondence to: Dr Janneke Berecki-Gisolf, Monash Injury Research Institute, Building 70 Clayton Campus, Monash University, Melbourne, Victoria 3800; e-mail: janneke.berecki-gisolf@monash.edu.au Submitted: September 2015; Revision requested: November 2015; Accepted: May 2016 The authors have stated they have no conflict of interest. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Aust NZ J Public Health. 2017; 41:85-91; doi: 10.1111/1753-6405.12568 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 85 © 2016 The Authors Berecki-Gisolf et al. Article 80 years, and fentanyl deaths have also T40.2: Other opioids (codeine, morphine), Pharmaceutical Benefits Scheme (PBS) increased; however, deaths from this cause T40.3: Methadone, and T40.4: Other synthetic The Australian Government subsidises are still uncommon and mainly (75%) involve narcotics (pethidine). For the purposes medicines through the PBS, which is available persons aged less than 47 years. Results of this study, drugs that are potentially to citizens and permanent residents. There from a study in the Australian state of Victoria available on prescription are considered is an out-of-pocket cost: for general patients report a nine-fold increase in oxycodone prescription drugs, even though these drugs this was set at $29.50 in 2006; this increased prescriptions between 2000 and 2009 that could also be obtained over the counter annually and reached $36.10 in 2013. For is more than matched by an increase in or illegitimately. Opium and heroin are not concession patients (e.g. pension card holders), deaths involving oxycodone. Codeine use available via prescription in Australia; however, this was $4.70 and $5.90, respectively. in Australia is difficult to capture because heroin (T40.1) or opium (T40.0) ICD-10-AM codeine can (to date) be obtained without codes occurring in the study sample, i.e. in Pharmaceutical prescription data prescription; furthermore, codeine is relatively combination with a T40.2, T40.3 or T40.4 code, Prescription drugs that were PBS-listed inexpensive and is frequently obtained were identified and reported in the study and subsidised through a PBS co-payment below the Australian Pharmaceutical Benefits results. Codeine combinations are (currently) were captured. Prescriptions purchased at a Scheme co-payment threshold. Codeine- available over the counter as well as on price below the PBS co-payment threshold related harm in Australia has increased: in prescription and are included in this study. were not included. All dispensings for 2000-09, accidental codeine overdose deaths Only incident admissions were selected: concessional beneficiaries are captured in increased from 1.8 to 5.1 deaths annually transfers within and between hospitals were these data as all drugs are priced above per million persons. Of accidental codeine excluded. Admissions that took place in the their co-payment, but dispensings for related deaths in 2000-13, the majority (84%) ED only, i.e. the patient received care in the general beneficiaries are not complete as involved multiple drug toxicity. ED only, throughout the recorded admission, some opioids are priced below the higher The sparse and inconsistent findings of previous were identified and excluded. Aggregate co-payment. PBS-listed medications were research are insufficient to support an informed tables were created by summing admissions provided with a unique PBS code specific to response to the problem of prescription opioid- by admission year, age group and gender. the drug type, form, dosage and pack size. related harm in Australia. Specifically, there is Intentional self-harm was identified as cases Summary PBS opioid prescription data were little information on age and gender differences with cause coded as (ICD-10-AM external requested from the Australian Government in opioid use in Australia, and none on the causes of morbidity and mortality: X60 Department of Human Services; availability relationship between the population being through to X69: Intentional self-poisoning). and interpretation of PBS data for research prescribed opioids and those most susceptible Other causes were grouped as ‘Unintentional is outlined in a recent publication by Mellish to opioid-related harm. poisoning’ (X40-X49), ‘Other unintentional’ et al (2015). The requested datasets were (X00-X39, X50-X59, W00-W46, W49-W81, More detailed understanding of the current stratified by age (<25 years, 25-44; 45-64 W83-W94, W99), ‘Other intentional self-harm’ trends in opioid prescribing and opioid and ≥65 years) and gender, and prescription (X70 - X84), ‘Assault’ (X85- X99, Y00- Y09), poisoning related hospital admissions and dispensing year. PBS items were grouped by ‘Poisoning - other and undetermined deaths in Australia by demographic groups is opioid type, based on anatomical therapeutic intent’ (Y10-Y19), and ‘Other – other and needed to adequately monitor the nature and chemical classification (ATC) coding. undetermined intent’ (Y20-Y36, Y40-Y89). extent of the public health consequences. Methadone and buprenorphine dispensings To address this need our study aimed to coded as drugs used in opioid dependence Death data (1) describe the population exposure to (N07BC02 and N07BC01, respectively) were prescription opioids, and (2) describe the Deaths in 2007-12 associated with opioid not included. Results were provided in terms frequency and distribution of opioid-related poisoning were determined from the Cause of the number of prescriptions as well as the hospital admissions and deaths, by age and of Death Unit Record file, provided to the number of persons receiving a prescription in gender. The analysis is limited to the State Victorian Injury Surveillance Unit by the each stratum. of Victoria, Australia, to provide population- Australian Coordinating Registry. These data Hospital admissions data based results for a defined population. are based on the death unit record files coded by the Australian Bureau of Statistics and Hospital admissions (2006 -14) related to are coded using the International Statistical opioid poisoning were determined from Methods Classification of Diseases and Related Health the Victorian Admitted Episodes Dataset Problems tenth revision (ICD-10). Death data The study was reviewed by the Monash (VAED). The VAED, in de-identified unit record is lagged and subject to revision pending University Human Research Ethics committee format, is provided annually to the Victorian coronial investigation: in this study, the and was granted exemption for ethical Injury Surveillance Unit by the Victorian death data from 2007-10 was final, data review. The research proposal satisfies section Department of Health and Human Services from 2011 was revised and data from 2012 5.1.22 of the National Statement on Ethical and is coded according to the International was preliminary. Deaths related to opioid Conduct in Human Research (‘institutions Statistical Classification of Diseases and poisoning were identified as those with may choose to exempt from ethical review Related Health Problems, Tenth Revision, an external cause and one of the following research that is negligible risk research, and Australian Modifications (ICD-10-AM). Hospital ICD10 codes in the record axis data (which involves the use of existing collections of data admissions related to opioid poisoning contain codes for all causes of death): or records that contain only non-identifiable were identified as those with one or more of T40.2: Other opioids (codeine, morphine), data about human beings’). the following ICD-10-AM diagnostic codes: 86 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning T40.3: Methadone, and T40.4: Other where α is the parameter estimate of year, overrepresentation of women among older synthetic narcotics (pethidine). Deaths were in the Poisson model. The analyses were persons, the female: male ratio in annual categorised according to year of registration, conducted using the PROC GENMOD dispensings per population for each age age group and gender. Intentional self-harm procedure in SAS V9.4. group is shown in Figure 1 (bottom). From was identified as cases with cause coded as 2006 to 2013, in every age group, the annual Prescription data, hospital data and death ‘Intentional self-poisoning’ (ICD-10 external rate was higher among women than men. data were not linked at a unit level: this study causes of morbidity and mortality: X60 is therefore a report of observed rates rather through to X69: Intentional self-poisoning). Prescription opioid dispensings: than a cohort with follow-up/outcomes study. drug types Residential population Of all captured opioid prescriptions, 29% Results The population of Victoria from 2006 to 2013 were for codeine with paracetamol, 27% was based on estimated resident population Prescription opioid dispensings: overall for oxycodone, 20% for tramadol, 10% data available from the Australian Bureau of for buprenorphine, 7.7% for morphine, Between 2006 and 2013, the annual number Statistics. Residential population data were 4.1% for fentanyl, 1.3% for oxycodone of PBS prescription opioid dispensings more stratified by year, age group and gender. with naloxone, 0.6% for methadone, 0.5% than doubled, from 1.64 million in 2006 to for hydromorphone, 0.05% for pethidine 3.32 million in 2013 (Table 1). During this Analysis hydrochloride and 0.003% for codeine with time, the residential population of Victoria Data were analysed using SAS 9.4. Population- aspirin. The trends in use of the seven most increased 13%, from 5.06 to 5.74 million. based dispensing rates were calculated commonly prescribed opioids, and relative Accounting for population growth, the annual by combining the pharmaceutical records use, are shown in Figure 2. In terms of number of prescription opioid dispensings with population data. Prescription opioid dispensings per 100 residents per year, each increased from 0.33 to 0.58 per resident: a dispensing was calculated in terms of annual of the opioid types increased in use with 78% increase between 2006 and 2013. prescriptions and annual prescriptions the exception of morphine. Between 2006 The annual prescription opioid dispensings per residential population. Where relevant, and 2013, there was a relative increase in per gender and age group are shown in the statistical significance of differences in oxycodone from 6.1 to 17.3 dispensings per Figure 1. Annual dispensings per population proportions was calculated with 2-tailed z-tests. 100 residents per year (an average annual increased statistically significantly from 0.38 The statistical significance of trends in annual increase of 16%), fentanyl from 0.5 to 2.3 (an in 2006 to 0.68 in 2013 (76% increase) among dispensings per population over time were average annual increase of 27%), codeine women, and from 0.26 in 2006 to 0.48 in modelled using Poisson models (proc genmod combinations from 11.8 to 15.7 (an average 2013 among men (81% increase). Annual in SAS), as trends in the annual number of annual increase of 5.2%), tramadol from 8.7 dispensings per population also increased dispensings, with the log of the annual Victorian to 10.1 (an average annual increase of 2.7%), in each of the age groups during this time residential population as offset. and oxycodone combinations emerged in period: from 0.02 to 0.07 in the age group of Deaths and hospital admissions were 2011. Buprenorphine dispensing increased under 24 years (204% increase); 0.24 to 0.42 combined with residential population data from 0.9 to 6.0 dispensings per 100 residents at ages 25-44 years (78% increase), 0.43 to and presented as frequencies and rates. per year (an average annual increase of 38%), 0.79 at ages 45-64 years (86% increase), 1.00 Population based rates were calculated as methadone dispensings increased from 0.22 to 1.41 at ages 65-84 years (42% increase), the number of events (deaths, admissions) to 0.31 (an average annual increase of 5.5%) and 1.61 to 3.07 at ages 85 years and over per residential population per year, and and morphine decreased by 6.4% per year on (90% increase). Of all prescription opioid presented as events per 1,000,000 person- average, from 4.1 to 2.5 dispensings per 100 dispensings in 2006, 8.1% were received by years, with 95% confidence intervals. Rates residents per year. persons age 85 years or over; in 2013 this were calculated per age group, gender and Age and gender trends in exposure to the increased to 10.6%. year, for hospital admissions and for deaths. strongest five opioids that were commonly To determine whether the high rates of Rates were also calculated separately for prescribed are shown in Tables 2-3. In opioid dispensings among women are due to each of the relevant ICD codes (T40.2, T40.3, terms of the number of persons receiving T40.4) and for intentional self-poisoning. For prescriptions at least once in the relevant Table 1: Trends in population size and opioid event counts below n=10, frequencies were year, the number of oxycodone, fentanyl and dispensings in Victoria, Australia 2006 -13. suppressed and rates were not calculated. a buprenorphine recipients increased between Year Residents Annual opioid Annual opioid Time trends in the rate of admissions were 2006 and 2013, both in the <65 years and the dispensings dispensings per modelled using Poisson models, as the b x 1000,000 x 1,000 population ≥65 years age groups. The most commonly annual number of events as a function of 2006 5.1 1,643 0.325 used strong opioid was oxycodone: in 2013, 2007 5.2 1,784 0.346 time in years (continuous), age group and 14% of Victorian residents aged ≥65 years 2008 5.3 1,921 0.365 gender, with the log of the annual Victorian filled at least one oxycodone prescription. 2009 5.4 2,022 0.376 residential population as offset. Trends in the The most pronounced gender difference in 2010 5.5 2,166 0.397 rate of deaths were not calculated because 2011 5.5 2,256 0.407 exposure to strong opioids was observed only five years of final data were available. 2012 5.6 2,937 0.521 for buprenorphine: by 2013 women 2013 5.7 3,322 0.579 Time trend results are presented as the were 2.2 times as likely as men to receive a: Based in Australian Bureau of Statistics data. modelled percentage change in rate per year, buprenorphine (1.44% vs. 0.65%; p<0.0002). b: Results are limited to prescribing according to the Pharmaceutical Benefits Scheme. calculated as: percentage change=[e -1]×100% 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 87 © 2016 The Authors Berecki-Gisolf et al. Article opioids among women was 2.15 that of Figure 1: Population based trends men. Admissions rates were highest in the in prescription opioid dispensings 25-44 year age groups, and lowest in the per resident per year, for 2006 -13 ≥85 year age groups, which is a strikingly in Victoria, Australia. Results are different pattern to that observed for opioid shown by gender (top), age group prescription use. There was a gradual increase (middle) and as male: female in the rate of hospital admissions, from 107 ratio, per age group (bottom). admissions per 1,000,000 person years in 2006 to 187 in 2014. Adjusted for age and gender, the rate increased by 6.8% per year. The rate of admissions related to ‘Other opioids – codeine, morphine’ increased by 6.5% per year; the rate of methadone related admissions increased by 11.7% per year, and the rate of ‘Other synthetic narcotics’ increased by 8.0% per year. The rate of opioid- related admissions that were considered to be due to self-harm increased by 8.4% per year (Table 4). Opioid-related harm: deaths Deaths associated with prescription opioid poisoning in Victoria, 2007-12, are summarised in Table 5. The most commonly occurring diagnosis was ‘other opioids – codeine, morphine’ (n=640, 73.1%), followed by methadone (n=254, 29.0%), followed by ‘other synthetic narcotics’ (n=76, 8.7%). Of all prescription opioid poisoning related deaths, 143 cases (16.3%) were considered to be intentional self-poisoning. Contrary to the patterns observed for hospital admissions, the rate of death related to prescription opioid poisoning was 1.90 times higher among men than women. This was true for each of the drug type categories. There were, however, no statistically significant gender differences in intentional self-poisoning death rates. The highest opioid poisoning related poisoning; 1102 (15%) poisoning of other Opioid-related harm: hospital death rate occurred the 25-44 year group; and undetermined intent; 154 (2%) other admissions the 45-64 year group had the highest rate unintentional, 31 (0.4%) other intentional Prescription opioid poisoning related of self-poisoning deaths. Cell counts of <10 self-harm and 14 admissions (0.2%) were due hospital admissions in Victoria 2006 -14 are deaths have been suppressed: for the ≥85 to assault. Among the admissions coded as summarised in Table 4: there were 7287 year age group, the death rate was therefore intentional self-poisoning, the distribution prescription-opioid related admissions below 11.1 deaths per 1,000,000 person-years of ICD-10-AM diagnostic codes was similar during this time period. The most commonly (the rate corresponding with 10 deaths in this to that in the overall prescription opioid assigned ICD-10-AM-coded diagnoses age group); which is less than a third of the poisoning admissions: in most cases (n=3435, (n=5874, 81%) were: ‘Other opioids: codeine; recorded death rate for the 25-44 year age 86%) the diagnostic code was: ‘Other opioids: morphine’ (T40.2), followed by: ‘Other group. Overall, the rate of opioid related deaths codeine; morphine’. Next most common synthetic narcotics: pethidine’ (ICD-10-AM: increased from 21 deaths per 1,000,000 person (n=628, 15%) were: ‘Other synthetic narcotics: T40.4) (n=1208, 17%), and: ‘Methadone’ years in 2007 to 28-29 in 2008-11, followed pethidine’ and ‘Methadone’ (n=169, 4%). In 18 (T40.3) (n=486, 7%). In 67 cases (1%), the by a drop to 26.5 in 2012, possibly due to the intentional self-poisoning cases (0.4%), the diagnostic code for heroin poisoning (T40.1) preliminary data for that year (which is to be diagnostic code for heroin poisoning (T40.1) was listed as well as a prescription opioid revised in the coming year). was listed as well as a prescription opioid poisoning code. poisoning code. Of all prescription opioid poisoning related Discussion During the study period (2006-14), the admissions in Victoria 2006 -13, 4050 (56%) hospital admission rate among women Our research findings in Victoria between admissions were considered to be intentional was 1.68 times higher than men (Table 4). 2006-07 and 2011-14 show a clear increase self-poisoning; 1764 (24%) unintentional The rate of intentional self-poisoning by in the frequency of prescription opioid 88 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning dispensings, particularly an increase in of the population being treated with poisoning deaths were more common among potent opioids such as oxycodone and prescription opioids does not fully correlate men. This pattern is not dissimilar to that fentanyl. Opioid-related harm also increased with prescription opioid poisoning rates. observed in the US, where drug overdose in this population. However, the problem deaths are more common among men. Overall, the rate of opioid dispensings in of prescription opioid dispensing and In the US, the gender gap is closing: death Victoria was greater among women than related harm in this population is not yet rates are increasing more rapidly among men in every age group; prescription opioid as severe as has been described in the women than men and prescription opioid hospital admissions were also more common US. Furthermore, the population-based overdose hospitalisations have also increased among women than men, but opioid relationship between exposure (high rates of prescription opioid dispensings) and adverse Figure 2: Trends in the seven most commonly prescribed opioids in dispensings per opioid type. outcomes (overdose, death) was not linear: there was an overrepresentation of adverse outcomes in younger people, relative to exposure. Few conclusions about the nature of the relationship between prescribing and harm can be drawn from the descriptive information presented, indicating a critical area where further studies are required. However, the increasing number of deaths associated with opioid poisoning and the recent steep increase in prescribing of certain types of opioids, makes this further research a necessity. Prescription opioid dispensing in Victoria increased by 78% over the period 2006 to 2013. Opioid use increased with increasing age, with highest rates of use among those aged ≥85 years. Opioid poisoning related hospital admission (2006 -14) and death rates (2007 -11) increased; opioid poisoning was most common among those aged 25-44 years. The overrepresentation of poisonings in younger people relative to dispensings suggests that in Victoria, increased numbers Table 2: Trends in the number of persons receiving prescriptions for the five most commonly prescribed strong opioids, by age group. Morphine Oxycodone Oxycodone combinations Fentanyl Buprenorphine Year Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) 2006 16,924 (0.4) 22,411 (3.3) 33,861 (0.8) 38,395 (5.7) - - 3,556 (0.1) 5,800 (0.9) 5,085 (0.1) 10,750 (1.6) 2007 15,090 (0.3) 21,609 (3.1) 43,194 (1.0) 50,850 (7.3) - - 5,096 (0.1) 8,435 (1.2) 7,702 (0.2) 21,432 (3.1) 2008 14,205 (0.3) 20,440 (2.9) 53,388 (1.2) 62,644 (8.8) - - 7,332 (0.2) 11,873 (1.7) 9,914 (0.2) 28,700 (4.0) 2009 13,558 (0.3) 20,027 (2.7) 64,380 (1.4) 75,055 (10.3) - - 9,110 (0.2) 14,587 (2.0) 11,066 (0.2) 32,599 (4.5) 2010 13,011 (0.3) 20,990 (2.8) 74,324 (1.6) 88,950 (11.8) - - 10,098 (0.2) 16,871 (2.2) 12,839 (0.3) 38,492 (5.1) 2011 12,617 (0.3) 22,932 (3.0) 83,003 (1.7) 98,627 (12.7) 221 (0.0) 252 (0.0) 10,389 (0.2) 17,971 (2.3) 13,532 (0.3) 41,691 (5.4) 2012 13,521 (0.3) 23,830 (3.0) 187,493 (3.9) 114,090 (14.1) 10,011 (0.2) 14,590 (1.8) 10,677 (0.2) 20,084 (2.5) 15,342 (0.3) 44,216 (5.5) 2013 13,225 (0.3) 22,972 (2.7) 234,805 (4.9) 117,994 (14.1) 23,729 (0.5) 31,575 (3.8) 9,900 (0.2) 19,279 (2.3) 15,193 (0.3) 45,067 (5.4) a: The number of persons who filled at least one prescription for the opioid type in the given year is presented as absolute number as well as percentage of the residential population in that age group Table 3: Trends in the number of persons receiving prescriptions for the five most commonly prescribed strong opioids, by gender. Morphine Oxycodone Oxycodone combinations Fentanyl Buprenorphine Year Women Men Women Men Women Men Women Men Women Men N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) 2006 21,018 (0.8) 18,317 (0.7) 40,774 (1.6) 31,482 (1.3) - - 5,209 (0.2) 4,147 (0.2) 10,950 (0.4) 4,885 (0.2) 2007 19,577 (0.8) 17,122 (0.7) 52,950 (2.0) 41,094 (1.6) - - 7,794 (0.3) 5,737 (0.2) 20,124 (0.8) 9,010 (0.4) 2008 18,596 (0.7) 16,049 (0.6) 65,939 (2.5) 50,093 (1.9) - - 11,515 (0.4) 7,690 (0.3) 26,887 (1.0) 11,727 (0.5) 2009 17,976 (0.7) 15,609 (0.6) 79,248 (2.9) 60,187 (2.3) - - 14,225 (0.5) 9,472 (0.4) 30,260 (1.1) 13,405 (0.5) 2010 18,224 (0.7) 15,777 (0.6) 93,405 (3.4) 69,869 (2.6) - - 16,401 (0.6) 10,568 (0.4) 35,816 (1.3) 15,515 (0.6) 2011 19,078 (0.7) 16,471 (0.6) 104,212 (3.7) 77,418 (2.8) 272 (0.0) 201 (0.0) 17,660 (0.6) 10,700 (0.4) 38,567 (1.4) 16,656 (0.6) 2012 19,943 (0.7) 17,408 (0.6) 164,933 (5.8) 136,650 (4.9) 14,515 (0.5) 10,086 (0.4) 19,509 (0.7) 11,252 (0.4) 41,437 (1.5) 18,121 (0.7) 2013 19,158 (0.7) 17,039 (0.6) 191,935 (6.6) 160,864 (5.7) 32,381 (1.1) 22,923 (0.8) 18,667 (0.6) 10,512 (0.4) 41,874 (1.4) 18,386 (0.6) a: The number of persons who filled at least one prescription for the opioid type in the given year is presented as absolute number as well as percentage of the population 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 89 © 2016 The Authors Berecki-Gisolf et al. Article more rapidly among women than men. risk factors, in terms of age and gender and Prescription opioid dispensing among For a better understanding of how trends in other sociodemographic factors, comorbidity, persons aged 85 years and over increased exposure impact on drug overdose and fatality and prescribing history including opioid by 90% between 2006 and 2013; particularly rates among men versus women, a population strength, type and prescription repeats. A oxycodone use. The deaths and hospital based data linkage study of prescription drug population-based data linkage study could admission analysis, however, did not show dispensing, hospital admissions and deaths is also provide insight into potential drug relatively high rates of opioid poisoning needed. This will provide insight into potential diversion pathways. among older persons. These findings support earlier reports showing that older persons are Table 4: Prescription opioid poisoning related hospital admissions in Victoria, 2006 -13. underrepresented among oxycodone-related Prescription opioid Diagnosis (ICD10) Intentional fatalities. The rapid increase in prescription Other opioids – Methadone Other synthetic – poisoning admissions, self-poisoning, opioid use among older persons, however, n=7,287 codeine, morphine (T40.3), n=486 narcotics (T40.4), n=4,050 warrants active monitoring of oxycodone- (T40.2), n=5,874 n=1,208 a b b b b b N Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) related hospitalisations and deaths, Gender particularly as older persons are likely to take M 2,684 111 (106–115) 83 (79–87) 11.2 (9.8–12.5) 21 (19–22) 52 (50–55) multiple medications and are therefore at risk F 4,603 186 (180–191) 156 (151–161) 8.7 (7.5–9.8) 29 (27–31) 112 (108–116) of multiple drug toxicity. This will facilitate Age group early identification of a trend following 0-24 1,429 90 (85–95) 72 (68–77) 4.7 (3.7–5.8) 16 (14–18) 59 (55–62) the opioid mortality observed in the US. 25-44 2,912 204 (196–211) 157 (150–163) 20.7 (18.3–23.0) 35 (32–38) 125 (119–131) Potential early interventions include revising 45-64 2,001 166 (159–174) 140 (134–147) 8.9 (7.2–10.6) 25 (22–27) 94 (88–99) 65-84 769 129 (120–138) 110 (102–119) – 20 (16–24) 31 (27–36) dosing guidelines and regulations regarding 85+ 176 195 (167–224) 153 (128–179) – 44 (31–58) 28 (17–39) potentially harmful drug combinations. Year Prescription opioid poisoning may not be a 2006 544 107 (98–117) 91 (83–99) 4.7 (2.8–6.6) 14 (11–18) 56 (49–62) direct consequence of increased prescription 2007 645 125 (115–135) 101 (93–110) 6.2 (4.1–8.4) 22 (18–26) 67 (60–74) 2008 694 132 (122–142) 107 (98–116) 8.9 (6.4–11.5) 20 (16–24) 72 (65–79) opioid use in the population: a more complex 2009 732 136 (126–146) 110 (101–119) 8.6 (6.1–11.0) 22 (18–26) 75 (68–82) conceptual model, including drug diversion, 2010 722 132 (123–142) 106 (97–114) 10.3 (7.6–12.9) 21 (18–25) 73 (65–80) tampering and other non-recommended 2011 833 150 (140–161) 118 (109–128) 10.5 (7.8–13.2) 29 (24–33) 81 (73–88) use, is required to fully understand how 2012 994 176 (166–187) 143 (133–152) 11.9 (9.0–14.7) 29 (24–33) 98 (90–106) prescribing trends and population-based 2013 1,029 179 (168–190) 144 (134–153) 13.4 (10.4–16.4) 30 (26–35) 105 (96–113) 2014 1,094 187 (176–198) 151 (141–161) 13.5 (10.5–16.5) 31 (27–36) 111 (102–119) exposure affect opioid poisoning trends. The Average % change +6.8% (5.2–8.4%) +6.5% (4.7–8.2%) +11.7% (7.4–16.2%) +8.0% (5.6–10.5%) +8.4% (6.4–10.5%) high proportion of hospital admissions that per year were coded as intentional self-poisoning a: Number of hospital admissions 2006 -14; transfers and in-hospital deaths excluded. suggests that mental health should also b: Rate of hospital admissions related to opioid poisoning, per 1,000,000 person-years 2006 -14. c: Calculated using a Poisson model of admissions per year, adjusted for age and gender; offset by population size. be taken into account. Future research to CI = confidence interval. provide more insight into the trends and potential for prevention and intervention Table 5: Prescription opioid poisoning related deaths in Victoria 2007–2012. could include analysis of coronial data to Prescription opioid poisoning Diagnosis (ICD10) Intentional establish comorbidity, substances involved Other opioids – Methadone Other synthetic – deaths, n=876 self–poisoning, and drug source, and circumstances of the codeine, morphine (T40.3), n=254 narcotics (T40.4), n=143 (T40.2), n=640 n=76 opioid poisoning. Data linkage studies of a b b b b b N Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) drug prescription data, hospital admissions Gender data and death data will further contribute to M 570 23.5 (21.5–25.4) 16.9 (15.3–18.5) 7.0 (6.0–8.1) 1.9 (1.4–2.5) 3.1 (2.4–3.8) knowledge of how drug prescribing trends F 306 12.4 (11.0–13.7) 9.3 (8.1–10.5) 3.4 (2.6–4.1) 1.2 (0.7–1.6) 2.7 (2.1–3.4) affect adverse consequences. Age group 0–24 60 3.8 (2.8–4.7) 2.5 (1.7–3.3) 1.4 (0.8–2.0) – This study’s main limitation relates to 25–44 501 35.1 (32.0–38.1) 24.9 (22.3–27.4) 11.8 (10.0–13.5) 2.4 (1.6–3.2) 3.3 (2.3–4.2) incompleteness in the prescription data. The 45–64 281 23.3 (20.6–26.1) 18.1 (15.7–20.5) 5.2 (3.9–6.4) 2.5 (1.6–3.4) 5.9 (4.5–7.3) following prescriptions are not captured: 65–84 25– 5.2 (3.4–7.0) 4.4 (2.7–6.0) 2.9 (1.5–4.2) 35 ‘private’ prescriptions; prescriptions 85+ <10 dispensed in some hospital settings; Year prescriptions obtained through the 2007 109 21.2 (17.2–25.1) 16.3 (12.8–19.8) 5.4 (3.4–7.4) – 2.9 (1.4–4.4) Repatriation Pharmaceutical Benefits Scheme 2008 151 28.7 (24.1–33.3) 24.2 (20.0–28.4) 7.2 (4.9–9.5) – 3.2 (1.7–4.8) (RPBS) and prescriptions that are below the 2009 154 28.7 (24.1–33.2) 22.7 (18.7–26.7) 6.9 (4.7–9.1) – 6.3 (4.2–8.5) 2010 155 28.4 (23.9–32.9) 19.0 (15.4–22.7) 8.8 (6.3–11.3) 2.7 (1.4–4.1) 4.8 (2.9–6.6) co-payment threshold and paid out of pocket. 2011 158 28.5 (24.1–33.0) 18.8 (15.2–22.4) 10.1 (7.5–12.8) 3.1 (1.6–4.5) 4.5 (2.7–6.3) These omissions may affect the described 2012 149 26.5 (22.2–30.7) 17.6 (14.1–21.0) 8.3 (6.0–10.7) 4.3 (2.6–6.0) 4.6 (2.8–6.4) use per age groups: drugs that are below the a: Number of deaths in 2007-12. co-payment threshold for regular patients b: Deaths per 1,000,000 person-years may appear in the PBS data for concession c: 2007-10 death data are final; 2011 death data has been revised, and the 2012 deaths are provided as preliminary data: these numbers are likely to change when the 2012 revised and final data are released. patients such as pension card holders. In a d: Exact number is not given to prevent recalculation of the low cell count for the ≥85 year age group. study of prescription opioid analgesics in CI = confidence interval. 90 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning 15. Centers for Disease Control and Prevention. Vital signs: Australia, 5.7% of oxycodone dispensings and older Australians; particularly oxycodone Overdoses of prescription opioid pain relievers and in 2008 were reported to be below the co- use among older persons has increased. other drugs among women-United States, 1999-2010. payment threshold, and 2.8% were private Hospital admissions (2006-14) and deaths MMWR Morb Mortal Wkly Rep. 2013;62(26):537-42. 16. Centers for Disease C Control and Prevention. Vital prescriptions. Of morphine dispensings, (2007-12) related to prescription opioid signs: Overdoses of prescription opioid pain relievers- 2.7% were under the co-payment threshold poisoning also increased, but these were United States, 1999-2008. MMWR Morb Mortal Wkly Rep. 2011;60(43):1487-92. and 3.4% were private prescriptions. Tramadol most common in the 25-44 year age group. 17. White AG, Birnbaum HG, Schiller M, Tang J, Katz and codeine had higher proportions of Furthermore, the majority of hospital NP. Analytic models to identify patients at risk for prescription opioid abuse. Am J Manag Care. non-co-payment dispensings: 12.8% and admissions were due to intentional self- 2009;15(12):897-906. 20.1%, respectively, and 4.3% and 6.7% were poisoning. The results emphasise the need 18. Hall AJ, Logan JE, Toblin RL, Kaplan JA, Kraner JC, private prescriptions. In another recent for more in-depth research on opioid-related Bixler D, et al. Patterns of abuse among unintentional pharmaceutical overdose fatalities. J AMA. study, the overall capture of prescription hospital admissions and deaths in Australia, 2008;300(22):2613-20. opioids in the PBS and RPBS in the period and a better understanding of the drivers 19. Hollingworth SA, Gray PD, Hall WD, Najman JM. Opioid analgesic prescribing in Australia: A focus on gender 1992 -11 were reported to be lower (68%); of prescribing and adverse consequences and age. Pharmacoepidemiol Drug Saf. 2015;24(6): 17% were reported to be obtained through of prescribing. Drug diversion, drug 628-36. 20. Blanch B, Pearson SA, Haber PS. An overview of the private prescriptions and 15% were below tampering as well as physical and mental patterns of prescription opioid use, costs and related the co-payment threshold (although these comorbidity need to be taken into account. harms in Australia. Br J Clin Pharmacol. 2014;78(5): are Australia-wide findings that may not In combination with monitoring of opioid 1159-66. 21. Leong M, Murnion B, Haber PS. Examination of opioid accurately reflect the situation in Victoria). prescribing and opioid poisoning, this will prescribing in Australia from 1992 to 2007. Intern Med Furthermore, over-the-counter codeine was facilitate early detection and prevention of a J. 2009;39(10):676-81. 22. Roxburgh A, Burns L, Drummer OH, Pilgrim J, Farrell not captured in this study: in Australia, 55.8% potential opioid epidemic. M, Degenhardt L. Trends in fentanyl prescriptions and of codeine packs sold in 2013 were over- fentanyl-related mortality in Australia. Drug Alcohol Rev. 2013;32(3):269-75. the-counter; for Victoria, this was 55.4%. References 23. Roxburgh A, Bruno R, Larance B, Burns L. Prescription of In other words, the current report under- opioid analgesics and related harms in Australia. Med J 1. Chen LH, Hedegaard H, Warner M. Drug-poisoning estimates actual opioid prescription filling. Aust. 2011;195(5):280-4. deaths involving opioid analgesics: United States, 24. Rintoul AC, Dobbin MD, Drummer OH, Ozanne-Smith 1999-2011. NCHS Data Brief. 2014;166:1-8. This inaccuracy is greater for younger persons J. Increasing deaths involving oxycodone, Victoria, 2. National Center for Health Statistics. Multiple Cause of than for older persons, as older persons Australia, 2000-09. Inj Prev. 2011;17(4):254-9. Death, 1999−2013 Query 2.2.15. In: CDC Wonder Online 25. Roxburgh A, Hall WD, Burns L, Pilgrim J, Saar E, Nielsen Database. Atlanta (GA): Centers for Disease Control and are more likely to receive prescriptions at S, et al. Trends and characteristics of accidental and Prevention; 2015. concessional rates (i.e. for older persons, intentional codeine overdose deaths in Australia. Med 3. National Institute on Drug Abuse. Overdose Death Rates J Aust. 2015;203(7):299. dispensings captured in the co-payment PBS [Internet]. Bethesda (MD): NIDA; 2015 [cited 2015 Sep 26. National Health and Medical Research Council. Chapter 7]. Available from: http://www.drugabuse.gov/related- records better reflect the total numbers of 5.1: Institutional responsibilities. In: National Statement topics/trends-statistics/overdose-death-rates dispensings). Age-differences in prescription on Ethical Conduct in Human Research (2007) (Updated 4. Rudd RA, Aleshire N, Zibbell JE, Gladden RM. Increases May 2015) [Internet]. Canberra (AUST): Government in drug and opioid deaths - United States, 2000-2014. filling are therefore likely to be overestimated of Australia; 2007 [cited 2015 Feb 4]. Available MMWR Morb Mortal Wkly Rep. 2016;64(50);1378-825 in this study. Another limitation is that this from: https://www.nhmrc.gov.au/book/chapter-5-1- 5. Centers for Disease Control and Prevention. Vital Signs: institutional-responsibilities Opioid Painkiller Prescribing. [Internet]. Atlanta (GA): study is based on separate data sources 27. Department of Health and Ageing. Pharmaceutical CDC; 2014 [2015 Sep 10]. Available from: http://www. that were not linked: access to unit level Benefits Scheme (PBS) [Internet]. Canberra (AUST): cdc.gov/vitalsigns/opioid-prescribing/index.html Government of Australia; 2015 [cited 2015 Jan 30]. prescription data and linkage with hospital 6. King NB, Fraser V, Boikos C, Richardson R, Harper S. Available from: http://www.pbs.gov.au/pbs/home Determinants of increased opioid-related mortality in and death data were not possible at the time 28. Mellish L, Karanges EA, Litchfield MJ, Schaffer AL, Blanch the United States and Canada, 1990-2013: A systematic of the study, but this will be the direction of B, Daniels BJ, et al. The Australian Pharmaceutical review. Am J Public Health. 2014;104(8):e32-42. Benefits Scheme data collection: A practical guide for 7. Boudreau D, Von Korff M, Rutter CM, Saunders K, Ray GT, future research in this area. researchers. BMC Res Notes . 2015;8:634. Sullivan MD, et al. Trends in long-term opioid therapy 29. Collaborating Centre for Drug Statistics Methodology. Victoria is the second-most populous state for chronic non-cancer pain. Pharmacoepidemiol Drug International Language for Drug Utilization Research Saf. 2009;18(12):1166-75. of Australia. In a recent study of opioid [Internet]. Geneva (CHE): World Health Organsiation; 8. Caudill-Slosberg MA, Schwartz LM, Woloshin S. Office 2014 [cited 2014 Jun 16]. Available from: http://www. dispensing in Australia, high-to-low ranking visits and analgesic prescriptions for musculoskeletal whocc.no/ pain in US: 1980 vs. 2000. Pain. 2004;109(3):514-9. of opioid dispensed through the PBS/RPBS 30. Australian Bureau of Statistics. Estimated Resident 9. Franklin GM, Mai J, Wickizer T, Turner JA, Fulton-Kehoe per state in 2013 shows Victoria as ranking Population (ERP) by Region, Age & Sex, 2001 to 2013 D, Grant L. Opioid dosing trends and mortality in 32 [Internet]. Canberra (AUST): ABS; 2015; [cited 2015 Washington State workers’ compensation, 1996-2002. fifth out of the eight jurisdictions. Victoria Jan 19]. Available from: http://stat.abs.gov.au/Index. Am J Ind Med. 2005;48(2):91-9. ranked fourth in oxycodone dispensings, aspx?DataSetCode=ABS_ERP_ASGS 10. Johnson H, Paulozzi L, Porucznik C, Mack K, Herter B, 31. Unick GJ, Rosenblum D, Mars S, Ciccarone D. fifth in tramadol dispensings and sixth Hal Johnson C, et al. Decline in drug overdose deaths Intertwined epidemics: national demographic trends after state policy changes - Florida, 2010-2012. MMWR in morphine dispensings. The results in hospitalizations for heroin- and opioid-related Morb Mortal Wkly Rep. 2014;63(26):569-74. overdoses, 1993-2009. PLoS One. 2013;8(2):e54496. presented in this study, although limited to 11. Paone D, Tuazon E, Kattan J, Nolan ML, O’Brien DB, 32. Mofizul Islam M, McRae IS, Mazumdar S, Taplin S, Dowell D, et al. Decrease in rate of opioid analgesic Victoria, do not misrepresent prescription McKetin R. Prescription opioid analgesics for pain overdose deaths - Staten Island, New York City, 2011- opioid dispensing in Australia. management in Australia: Twenty years of dispensing. 2013. MMWR Morb Mortal Wkly Rep. 2015;64(18):491-4. Intern Med J. 2015. doi: 10.1111/imj.12966. 12. Rudd RA, Paulozzi LJ, Bauer MJ, Burleson RW, Carlson 33. Gisev N, Nielsen S, Cama E, Larance B, Bruno R, RE, Dao D, et al. Increases in heroin overdose deaths - Degenhardt L. An ecological study of the extent and 28 States, 2010 to 2012. MMWR Morb Mortal Wkly Rep. Conclusion and implications factors associated with the use of prescription and over- 2014;63(39):849-54. the-counter codeine in Australia. Eur J Clin Pharmacol. 13. Compton WM, Jones CM, Baldwin GT. Relationship Overall prescription opioid consumption 2016;72(4):469-94. between nonmedical prescription-opioid use and heroin use. N Engl J Med. 2016;374(2):154-63. has increased rapidly in Victoria in 2006-13. 14. Rudd RA, Aleshire N, Zibbell JE, Gladden RM. Increases Increase in prescription opioid dispensing in drug and opioid overdose deaths - United States, has been most pronounced among women 2000-2014. MMWR Morb Mortal Wkly Rep. 2016;64(50- 51):1378-82. 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 91 © 2016 The Authors http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Prescription opioid dispensing and prescription opioid poisoning: Population data from Victoria, Australia 2006 to 2013

Download PDF
7 pages

Loading next page...
 
/lp/wiley/prescription-opioid-dispensing-and-prescription-opioid-poisoning-xCqMiQ0BvJ

References (47)

Publisher
Wiley
Copyright
© 2017 Public Health Association of Australia
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/1753-6405.12568
pmid
27624336
Publisher site
See Article on Publisher Site

Abstract

been referred to as epidemic in the Objective: To describe recent trends in opioid prescribing and prescription opioid poisoning PUnited States (US). Between 2001 resulting in hospitalisation or death in Victoria, Australia. and 2014 there was a three-fold increase in Method: This is a population-based ecological study of residents of Victoria, 2006 -14. deaths from this cause, plateauing in 2011 at 2-4 Australian Bureau of Statistics residential population data were combined with Pharmaceutical about 17,000 deaths annually. US health Benefits Scheme (PBS) opioid prescription data, Victorian Admitted Episodes Data (VAED) and agencies have recognised the national cause of death data. increase in opioid prescriptions as a key driver Results: Annual opioid dispensings increased by 78% in 2006 -13, from 0.33 to 0.58 per of the increase in prescription drug overdose population. Opioid use increased with age: in 2013, 14% of Victorian residents aged ≥65 years deaths; particularly the prescribing of long- 5-9 filled at least one oxycodone prescription. In 2006 -14, prescription opioid related hospital term opioids for chronic non-cancer pain. admissions increased by 6.8% per year, from 107 to 187 /1,000,000 person-years; 56% were A number of US states have implemented due to intentional self-poisoning. Annual deaths increased from 21 to 28 /1,000,000 persons, in a series of policy responses focused on 2007 -11. Admissions and deaths peaked at 25-44 years. restriction of prescription opioid supply and Conclusions: Although both opioid prescribing and poisoning have increased, there is increased availability of the opioid antagonist 10,11 discrepancy between the exposed group (dispensings increased with age) and those with drug naloxone. While there has been adverse consequences (rates peaked at ages 25-44 years). a plateau in the non-fentanyl prescription Implications: A better understanding is needed of drivers of prescribing and adverse opioid epidemic curve that has coincided consequences. Together with monitoring of prescribing and poisoning, this will facilitate early with the constraints on prescription supply, detection and prevention of a public health problem. heroin deaths in the US since 2011 have Key words: prescription opioids, drug overdose, pharmacoepidemiology tripled, with consequent ongoing escalation of the overall opioid-related public health 2,3,12,13 problem. Fentanyl can be obtained currently have a death rate that is increasing appears somewhat resistant to intervention. on prescription or manufactured illicitly: more quickly than among men. The highest There are early indications that Australia has on the death certificate, deaths involving rate of deaths in both sexes is in the 45-55 age a prescription opioid problem. Annual opioid prescription fentanyl cannot be distinguished group. Individuals at highest risk of opioid analgesic use has increased in Australia 2002- from deaths involving illicitly manufactured related harm are those being prescribed long 2009, from 13 to 16 defined daily dosages fentanyl. The availability of illicitly term, high Morphine Equivalent Doses and (DDD) per 1,000 population per day. A study manufactured fentanyl has increased in the those obtaining concurrent prescriptions of opioid prescribing and adverse outcomes 14 17,18 US. In the US, the relationship between from different sources. There are three key in Australia reported a 15-fold increase in opioid prescriptions and opioid deaths has lessons to be learned from the US experience. opioid prescribing between 1992 and 2012, been explored in some detail. Regular opioid First, there is a substantial period between the but with an apparently small effect on opioid use in the US is more commonly reported beginning of the epidemic and the point at related harm. Oxycodone and fentanyl among women than men, with women more which a national response can be mobilised. prescribing have both increased since 2000, likely to use opioids chronically at higher Second, the increase in opioid deaths appears but unlike the US, the rate of increased doses and to increase their use with age. to be related to an increase in the amount prescribing was not shown to match the rate 21-23 Deaths are more frequent among men than of opioids prescribed, both at a population of increase in deaths. Fentanyl prescribing women, although women are hospitalised level and in relation to the group for whom in Australia has increased in recent years, for opioid overdose more frequently and prescriptions are intended. Third, the problem particularly among persons aged over 1. Monash Injury Research Institute, Monash University, Victoria 2. Institute for Safety, Compensation and Recovery Research, Monash University, Victoria 3. Harvard Injury Control Research Center, Harvard School of Population Health, Massachusetts, USA Correspondence to: Dr Janneke Berecki-Gisolf, Monash Injury Research Institute, Building 70 Clayton Campus, Monash University, Melbourne, Victoria 3800; e-mail: janneke.berecki-gisolf@monash.edu.au Submitted: September 2015; Revision requested: November 2015; Accepted: May 2016 The authors have stated they have no conflict of interest. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Aust NZ J Public Health. 2017; 41:85-91; doi: 10.1111/1753-6405.12568 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 85 © 2016 The Authors Berecki-Gisolf et al. Article 80 years, and fentanyl deaths have also T40.2: Other opioids (codeine, morphine), Pharmaceutical Benefits Scheme (PBS) increased; however, deaths from this cause T40.3: Methadone, and T40.4: Other synthetic The Australian Government subsidises are still uncommon and mainly (75%) involve narcotics (pethidine). For the purposes medicines through the PBS, which is available persons aged less than 47 years. Results of this study, drugs that are potentially to citizens and permanent residents. There from a study in the Australian state of Victoria available on prescription are considered is an out-of-pocket cost: for general patients report a nine-fold increase in oxycodone prescription drugs, even though these drugs this was set at $29.50 in 2006; this increased prescriptions between 2000 and 2009 that could also be obtained over the counter annually and reached $36.10 in 2013. For is more than matched by an increase in or illegitimately. Opium and heroin are not concession patients (e.g. pension card holders), deaths involving oxycodone. Codeine use available via prescription in Australia; however, this was $4.70 and $5.90, respectively. in Australia is difficult to capture because heroin (T40.1) or opium (T40.0) ICD-10-AM codeine can (to date) be obtained without codes occurring in the study sample, i.e. in Pharmaceutical prescription data prescription; furthermore, codeine is relatively combination with a T40.2, T40.3 or T40.4 code, Prescription drugs that were PBS-listed inexpensive and is frequently obtained were identified and reported in the study and subsidised through a PBS co-payment below the Australian Pharmaceutical Benefits results. Codeine combinations are (currently) were captured. Prescriptions purchased at a Scheme co-payment threshold. Codeine- available over the counter as well as on price below the PBS co-payment threshold related harm in Australia has increased: in prescription and are included in this study. were not included. All dispensings for 2000-09, accidental codeine overdose deaths Only incident admissions were selected: concessional beneficiaries are captured in increased from 1.8 to 5.1 deaths annually transfers within and between hospitals were these data as all drugs are priced above per million persons. Of accidental codeine excluded. Admissions that took place in the their co-payment, but dispensings for related deaths in 2000-13, the majority (84%) ED only, i.e. the patient received care in the general beneficiaries are not complete as involved multiple drug toxicity. ED only, throughout the recorded admission, some opioids are priced below the higher The sparse and inconsistent findings of previous were identified and excluded. Aggregate co-payment. PBS-listed medications were research are insufficient to support an informed tables were created by summing admissions provided with a unique PBS code specific to response to the problem of prescription opioid- by admission year, age group and gender. the drug type, form, dosage and pack size. related harm in Australia. Specifically, there is Intentional self-harm was identified as cases Summary PBS opioid prescription data were little information on age and gender differences with cause coded as (ICD-10-AM external requested from the Australian Government in opioid use in Australia, and none on the causes of morbidity and mortality: X60 Department of Human Services; availability relationship between the population being through to X69: Intentional self-poisoning). and interpretation of PBS data for research prescribed opioids and those most susceptible Other causes were grouped as ‘Unintentional is outlined in a recent publication by Mellish to opioid-related harm. poisoning’ (X40-X49), ‘Other unintentional’ et al (2015). The requested datasets were (X00-X39, X50-X59, W00-W46, W49-W81, More detailed understanding of the current stratified by age (<25 years, 25-44; 45-64 W83-W94, W99), ‘Other intentional self-harm’ trends in opioid prescribing and opioid and ≥65 years) and gender, and prescription (X70 - X84), ‘Assault’ (X85- X99, Y00- Y09), poisoning related hospital admissions and dispensing year. PBS items were grouped by ‘Poisoning - other and undetermined deaths in Australia by demographic groups is opioid type, based on anatomical therapeutic intent’ (Y10-Y19), and ‘Other – other and needed to adequately monitor the nature and chemical classification (ATC) coding. undetermined intent’ (Y20-Y36, Y40-Y89). extent of the public health consequences. Methadone and buprenorphine dispensings To address this need our study aimed to coded as drugs used in opioid dependence Death data (1) describe the population exposure to (N07BC02 and N07BC01, respectively) were prescription opioids, and (2) describe the Deaths in 2007-12 associated with opioid not included. Results were provided in terms frequency and distribution of opioid-related poisoning were determined from the Cause of the number of prescriptions as well as the hospital admissions and deaths, by age and of Death Unit Record file, provided to the number of persons receiving a prescription in gender. The analysis is limited to the State Victorian Injury Surveillance Unit by the each stratum. of Victoria, Australia, to provide population- Australian Coordinating Registry. These data Hospital admissions data based results for a defined population. are based on the death unit record files coded by the Australian Bureau of Statistics and Hospital admissions (2006 -14) related to are coded using the International Statistical opioid poisoning were determined from Methods Classification of Diseases and Related Health the Victorian Admitted Episodes Dataset Problems tenth revision (ICD-10). Death data The study was reviewed by the Monash (VAED). The VAED, in de-identified unit record is lagged and subject to revision pending University Human Research Ethics committee format, is provided annually to the Victorian coronial investigation: in this study, the and was granted exemption for ethical Injury Surveillance Unit by the Victorian death data from 2007-10 was final, data review. The research proposal satisfies section Department of Health and Human Services from 2011 was revised and data from 2012 5.1.22 of the National Statement on Ethical and is coded according to the International was preliminary. Deaths related to opioid Conduct in Human Research (‘institutions Statistical Classification of Diseases and poisoning were identified as those with may choose to exempt from ethical review Related Health Problems, Tenth Revision, an external cause and one of the following research that is negligible risk research, and Australian Modifications (ICD-10-AM). Hospital ICD10 codes in the record axis data (which involves the use of existing collections of data admissions related to opioid poisoning contain codes for all causes of death): or records that contain only non-identifiable were identified as those with one or more of T40.2: Other opioids (codeine, morphine), data about human beings’). the following ICD-10-AM diagnostic codes: 86 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning T40.3: Methadone, and T40.4: Other where α is the parameter estimate of year, overrepresentation of women among older synthetic narcotics (pethidine). Deaths were in the Poisson model. The analyses were persons, the female: male ratio in annual categorised according to year of registration, conducted using the PROC GENMOD dispensings per population for each age age group and gender. Intentional self-harm procedure in SAS V9.4. group is shown in Figure 1 (bottom). From was identified as cases with cause coded as 2006 to 2013, in every age group, the annual Prescription data, hospital data and death ‘Intentional self-poisoning’ (ICD-10 external rate was higher among women than men. data were not linked at a unit level: this study causes of morbidity and mortality: X60 is therefore a report of observed rates rather through to X69: Intentional self-poisoning). Prescription opioid dispensings: than a cohort with follow-up/outcomes study. drug types Residential population Of all captured opioid prescriptions, 29% Results The population of Victoria from 2006 to 2013 were for codeine with paracetamol, 27% was based on estimated resident population Prescription opioid dispensings: overall for oxycodone, 20% for tramadol, 10% data available from the Australian Bureau of for buprenorphine, 7.7% for morphine, Between 2006 and 2013, the annual number Statistics. Residential population data were 4.1% for fentanyl, 1.3% for oxycodone of PBS prescription opioid dispensings more stratified by year, age group and gender. with naloxone, 0.6% for methadone, 0.5% than doubled, from 1.64 million in 2006 to for hydromorphone, 0.05% for pethidine 3.32 million in 2013 (Table 1). During this Analysis hydrochloride and 0.003% for codeine with time, the residential population of Victoria Data were analysed using SAS 9.4. Population- aspirin. The trends in use of the seven most increased 13%, from 5.06 to 5.74 million. based dispensing rates were calculated commonly prescribed opioids, and relative Accounting for population growth, the annual by combining the pharmaceutical records use, are shown in Figure 2. In terms of number of prescription opioid dispensings with population data. Prescription opioid dispensings per 100 residents per year, each increased from 0.33 to 0.58 per resident: a dispensing was calculated in terms of annual of the opioid types increased in use with 78% increase between 2006 and 2013. prescriptions and annual prescriptions the exception of morphine. Between 2006 The annual prescription opioid dispensings per residential population. Where relevant, and 2013, there was a relative increase in per gender and age group are shown in the statistical significance of differences in oxycodone from 6.1 to 17.3 dispensings per Figure 1. Annual dispensings per population proportions was calculated with 2-tailed z-tests. 100 residents per year (an average annual increased statistically significantly from 0.38 The statistical significance of trends in annual increase of 16%), fentanyl from 0.5 to 2.3 (an in 2006 to 0.68 in 2013 (76% increase) among dispensings per population over time were average annual increase of 27%), codeine women, and from 0.26 in 2006 to 0.48 in modelled using Poisson models (proc genmod combinations from 11.8 to 15.7 (an average 2013 among men (81% increase). Annual in SAS), as trends in the annual number of annual increase of 5.2%), tramadol from 8.7 dispensings per population also increased dispensings, with the log of the annual Victorian to 10.1 (an average annual increase of 2.7%), in each of the age groups during this time residential population as offset. and oxycodone combinations emerged in period: from 0.02 to 0.07 in the age group of Deaths and hospital admissions were 2011. Buprenorphine dispensing increased under 24 years (204% increase); 0.24 to 0.42 combined with residential population data from 0.9 to 6.0 dispensings per 100 residents at ages 25-44 years (78% increase), 0.43 to and presented as frequencies and rates. per year (an average annual increase of 38%), 0.79 at ages 45-64 years (86% increase), 1.00 Population based rates were calculated as methadone dispensings increased from 0.22 to 1.41 at ages 65-84 years (42% increase), the number of events (deaths, admissions) to 0.31 (an average annual increase of 5.5%) and 1.61 to 3.07 at ages 85 years and over per residential population per year, and and morphine decreased by 6.4% per year on (90% increase). Of all prescription opioid presented as events per 1,000,000 person- average, from 4.1 to 2.5 dispensings per 100 dispensings in 2006, 8.1% were received by years, with 95% confidence intervals. Rates residents per year. persons age 85 years or over; in 2013 this were calculated per age group, gender and Age and gender trends in exposure to the increased to 10.6%. year, for hospital admissions and for deaths. strongest five opioids that were commonly To determine whether the high rates of Rates were also calculated separately for prescribed are shown in Tables 2-3. In opioid dispensings among women are due to each of the relevant ICD codes (T40.2, T40.3, terms of the number of persons receiving T40.4) and for intentional self-poisoning. For prescriptions at least once in the relevant Table 1: Trends in population size and opioid event counts below n=10, frequencies were year, the number of oxycodone, fentanyl and dispensings in Victoria, Australia 2006 -13. suppressed and rates were not calculated. a buprenorphine recipients increased between Year Residents Annual opioid Annual opioid Time trends in the rate of admissions were 2006 and 2013, both in the <65 years and the dispensings dispensings per modelled using Poisson models, as the b x 1000,000 x 1,000 population ≥65 years age groups. The most commonly annual number of events as a function of 2006 5.1 1,643 0.325 used strong opioid was oxycodone: in 2013, 2007 5.2 1,784 0.346 time in years (continuous), age group and 14% of Victorian residents aged ≥65 years 2008 5.3 1,921 0.365 gender, with the log of the annual Victorian filled at least one oxycodone prescription. 2009 5.4 2,022 0.376 residential population as offset. Trends in the The most pronounced gender difference in 2010 5.5 2,166 0.397 rate of deaths were not calculated because 2011 5.5 2,256 0.407 exposure to strong opioids was observed only five years of final data were available. 2012 5.6 2,937 0.521 for buprenorphine: by 2013 women 2013 5.7 3,322 0.579 Time trend results are presented as the were 2.2 times as likely as men to receive a: Based in Australian Bureau of Statistics data. modelled percentage change in rate per year, buprenorphine (1.44% vs. 0.65%; p<0.0002). b: Results are limited to prescribing according to the Pharmaceutical Benefits Scheme. calculated as: percentage change=[e -1]×100% 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 87 © 2016 The Authors Berecki-Gisolf et al. Article opioids among women was 2.15 that of Figure 1: Population based trends men. Admissions rates were highest in the in prescription opioid dispensings 25-44 year age groups, and lowest in the per resident per year, for 2006 -13 ≥85 year age groups, which is a strikingly in Victoria, Australia. Results are different pattern to that observed for opioid shown by gender (top), age group prescription use. There was a gradual increase (middle) and as male: female in the rate of hospital admissions, from 107 ratio, per age group (bottom). admissions per 1,000,000 person years in 2006 to 187 in 2014. Adjusted for age and gender, the rate increased by 6.8% per year. The rate of admissions related to ‘Other opioids – codeine, morphine’ increased by 6.5% per year; the rate of methadone related admissions increased by 11.7% per year, and the rate of ‘Other synthetic narcotics’ increased by 8.0% per year. The rate of opioid- related admissions that were considered to be due to self-harm increased by 8.4% per year (Table 4). Opioid-related harm: deaths Deaths associated with prescription opioid poisoning in Victoria, 2007-12, are summarised in Table 5. The most commonly occurring diagnosis was ‘other opioids – codeine, morphine’ (n=640, 73.1%), followed by methadone (n=254, 29.0%), followed by ‘other synthetic narcotics’ (n=76, 8.7%). Of all prescription opioid poisoning related deaths, 143 cases (16.3%) were considered to be intentional self-poisoning. Contrary to the patterns observed for hospital admissions, the rate of death related to prescription opioid poisoning was 1.90 times higher among men than women. This was true for each of the drug type categories. There were, however, no statistically significant gender differences in intentional self-poisoning death rates. The highest opioid poisoning related poisoning; 1102 (15%) poisoning of other Opioid-related harm: hospital death rate occurred the 25-44 year group; and undetermined intent; 154 (2%) other admissions the 45-64 year group had the highest rate unintentional, 31 (0.4%) other intentional Prescription opioid poisoning related of self-poisoning deaths. Cell counts of <10 self-harm and 14 admissions (0.2%) were due hospital admissions in Victoria 2006 -14 are deaths have been suppressed: for the ≥85 to assault. Among the admissions coded as summarised in Table 4: there were 7287 year age group, the death rate was therefore intentional self-poisoning, the distribution prescription-opioid related admissions below 11.1 deaths per 1,000,000 person-years of ICD-10-AM diagnostic codes was similar during this time period. The most commonly (the rate corresponding with 10 deaths in this to that in the overall prescription opioid assigned ICD-10-AM-coded diagnoses age group); which is less than a third of the poisoning admissions: in most cases (n=3435, (n=5874, 81%) were: ‘Other opioids: codeine; recorded death rate for the 25-44 year age 86%) the diagnostic code was: ‘Other opioids: morphine’ (T40.2), followed by: ‘Other group. Overall, the rate of opioid related deaths codeine; morphine’. Next most common synthetic narcotics: pethidine’ (ICD-10-AM: increased from 21 deaths per 1,000,000 person (n=628, 15%) were: ‘Other synthetic narcotics: T40.4) (n=1208, 17%), and: ‘Methadone’ years in 2007 to 28-29 in 2008-11, followed pethidine’ and ‘Methadone’ (n=169, 4%). In 18 (T40.3) (n=486, 7%). In 67 cases (1%), the by a drop to 26.5 in 2012, possibly due to the intentional self-poisoning cases (0.4%), the diagnostic code for heroin poisoning (T40.1) preliminary data for that year (which is to be diagnostic code for heroin poisoning (T40.1) was listed as well as a prescription opioid revised in the coming year). was listed as well as a prescription opioid poisoning code. poisoning code. Of all prescription opioid poisoning related Discussion During the study period (2006-14), the admissions in Victoria 2006 -13, 4050 (56%) hospital admission rate among women Our research findings in Victoria between admissions were considered to be intentional was 1.68 times higher than men (Table 4). 2006-07 and 2011-14 show a clear increase self-poisoning; 1764 (24%) unintentional The rate of intentional self-poisoning by in the frequency of prescription opioid 88 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning dispensings, particularly an increase in of the population being treated with poisoning deaths were more common among potent opioids such as oxycodone and prescription opioids does not fully correlate men. This pattern is not dissimilar to that fentanyl. Opioid-related harm also increased with prescription opioid poisoning rates. observed in the US, where drug overdose in this population. However, the problem deaths are more common among men. Overall, the rate of opioid dispensings in of prescription opioid dispensing and In the US, the gender gap is closing: death Victoria was greater among women than related harm in this population is not yet rates are increasing more rapidly among men in every age group; prescription opioid as severe as has been described in the women than men and prescription opioid hospital admissions were also more common US. Furthermore, the population-based overdose hospitalisations have also increased among women than men, but opioid relationship between exposure (high rates of prescription opioid dispensings) and adverse Figure 2: Trends in the seven most commonly prescribed opioids in dispensings per opioid type. outcomes (overdose, death) was not linear: there was an overrepresentation of adverse outcomes in younger people, relative to exposure. Few conclusions about the nature of the relationship between prescribing and harm can be drawn from the descriptive information presented, indicating a critical area where further studies are required. However, the increasing number of deaths associated with opioid poisoning and the recent steep increase in prescribing of certain types of opioids, makes this further research a necessity. Prescription opioid dispensing in Victoria increased by 78% over the period 2006 to 2013. Opioid use increased with increasing age, with highest rates of use among those aged ≥85 years. Opioid poisoning related hospital admission (2006 -14) and death rates (2007 -11) increased; opioid poisoning was most common among those aged 25-44 years. The overrepresentation of poisonings in younger people relative to dispensings suggests that in Victoria, increased numbers Table 2: Trends in the number of persons receiving prescriptions for the five most commonly prescribed strong opioids, by age group. Morphine Oxycodone Oxycodone combinations Fentanyl Buprenorphine Year Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged Persons aged <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) 2006 16,924 (0.4) 22,411 (3.3) 33,861 (0.8) 38,395 (5.7) - - 3,556 (0.1) 5,800 (0.9) 5,085 (0.1) 10,750 (1.6) 2007 15,090 (0.3) 21,609 (3.1) 43,194 (1.0) 50,850 (7.3) - - 5,096 (0.1) 8,435 (1.2) 7,702 (0.2) 21,432 (3.1) 2008 14,205 (0.3) 20,440 (2.9) 53,388 (1.2) 62,644 (8.8) - - 7,332 (0.2) 11,873 (1.7) 9,914 (0.2) 28,700 (4.0) 2009 13,558 (0.3) 20,027 (2.7) 64,380 (1.4) 75,055 (10.3) - - 9,110 (0.2) 14,587 (2.0) 11,066 (0.2) 32,599 (4.5) 2010 13,011 (0.3) 20,990 (2.8) 74,324 (1.6) 88,950 (11.8) - - 10,098 (0.2) 16,871 (2.2) 12,839 (0.3) 38,492 (5.1) 2011 12,617 (0.3) 22,932 (3.0) 83,003 (1.7) 98,627 (12.7) 221 (0.0) 252 (0.0) 10,389 (0.2) 17,971 (2.3) 13,532 (0.3) 41,691 (5.4) 2012 13,521 (0.3) 23,830 (3.0) 187,493 (3.9) 114,090 (14.1) 10,011 (0.2) 14,590 (1.8) 10,677 (0.2) 20,084 (2.5) 15,342 (0.3) 44,216 (5.5) 2013 13,225 (0.3) 22,972 (2.7) 234,805 (4.9) 117,994 (14.1) 23,729 (0.5) 31,575 (3.8) 9,900 (0.2) 19,279 (2.3) 15,193 (0.3) 45,067 (5.4) a: The number of persons who filled at least one prescription for the opioid type in the given year is presented as absolute number as well as percentage of the residential population in that age group Table 3: Trends in the number of persons receiving prescriptions for the five most commonly prescribed strong opioids, by gender. Morphine Oxycodone Oxycodone combinations Fentanyl Buprenorphine Year Women Men Women Men Women Men Women Men Women Men N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) N (%) 2006 21,018 (0.8) 18,317 (0.7) 40,774 (1.6) 31,482 (1.3) - - 5,209 (0.2) 4,147 (0.2) 10,950 (0.4) 4,885 (0.2) 2007 19,577 (0.8) 17,122 (0.7) 52,950 (2.0) 41,094 (1.6) - - 7,794 (0.3) 5,737 (0.2) 20,124 (0.8) 9,010 (0.4) 2008 18,596 (0.7) 16,049 (0.6) 65,939 (2.5) 50,093 (1.9) - - 11,515 (0.4) 7,690 (0.3) 26,887 (1.0) 11,727 (0.5) 2009 17,976 (0.7) 15,609 (0.6) 79,248 (2.9) 60,187 (2.3) - - 14,225 (0.5) 9,472 (0.4) 30,260 (1.1) 13,405 (0.5) 2010 18,224 (0.7) 15,777 (0.6) 93,405 (3.4) 69,869 (2.6) - - 16,401 (0.6) 10,568 (0.4) 35,816 (1.3) 15,515 (0.6) 2011 19,078 (0.7) 16,471 (0.6) 104,212 (3.7) 77,418 (2.8) 272 (0.0) 201 (0.0) 17,660 (0.6) 10,700 (0.4) 38,567 (1.4) 16,656 (0.6) 2012 19,943 (0.7) 17,408 (0.6) 164,933 (5.8) 136,650 (4.9) 14,515 (0.5) 10,086 (0.4) 19,509 (0.7) 11,252 (0.4) 41,437 (1.5) 18,121 (0.7) 2013 19,158 (0.7) 17,039 (0.6) 191,935 (6.6) 160,864 (5.7) 32,381 (1.1) 22,923 (0.8) 18,667 (0.6) 10,512 (0.4) 41,874 (1.4) 18,386 (0.6) a: The number of persons who filled at least one prescription for the opioid type in the given year is presented as absolute number as well as percentage of the population 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 89 © 2016 The Authors Berecki-Gisolf et al. Article more rapidly among women than men. risk factors, in terms of age and gender and Prescription opioid dispensing among For a better understanding of how trends in other sociodemographic factors, comorbidity, persons aged 85 years and over increased exposure impact on drug overdose and fatality and prescribing history including opioid by 90% between 2006 and 2013; particularly rates among men versus women, a population strength, type and prescription repeats. A oxycodone use. The deaths and hospital based data linkage study of prescription drug population-based data linkage study could admission analysis, however, did not show dispensing, hospital admissions and deaths is also provide insight into potential drug relatively high rates of opioid poisoning needed. This will provide insight into potential diversion pathways. among older persons. These findings support earlier reports showing that older persons are Table 4: Prescription opioid poisoning related hospital admissions in Victoria, 2006 -13. underrepresented among oxycodone-related Prescription opioid Diagnosis (ICD10) Intentional fatalities. The rapid increase in prescription Other opioids – Methadone Other synthetic – poisoning admissions, self-poisoning, opioid use among older persons, however, n=7,287 codeine, morphine (T40.3), n=486 narcotics (T40.4), n=4,050 warrants active monitoring of oxycodone- (T40.2), n=5,874 n=1,208 a b b b b b N Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) related hospitalisations and deaths, Gender particularly as older persons are likely to take M 2,684 111 (106–115) 83 (79–87) 11.2 (9.8–12.5) 21 (19–22) 52 (50–55) multiple medications and are therefore at risk F 4,603 186 (180–191) 156 (151–161) 8.7 (7.5–9.8) 29 (27–31) 112 (108–116) of multiple drug toxicity. This will facilitate Age group early identification of a trend following 0-24 1,429 90 (85–95) 72 (68–77) 4.7 (3.7–5.8) 16 (14–18) 59 (55–62) the opioid mortality observed in the US. 25-44 2,912 204 (196–211) 157 (150–163) 20.7 (18.3–23.0) 35 (32–38) 125 (119–131) Potential early interventions include revising 45-64 2,001 166 (159–174) 140 (134–147) 8.9 (7.2–10.6) 25 (22–27) 94 (88–99) 65-84 769 129 (120–138) 110 (102–119) – 20 (16–24) 31 (27–36) dosing guidelines and regulations regarding 85+ 176 195 (167–224) 153 (128–179) – 44 (31–58) 28 (17–39) potentially harmful drug combinations. Year Prescription opioid poisoning may not be a 2006 544 107 (98–117) 91 (83–99) 4.7 (2.8–6.6) 14 (11–18) 56 (49–62) direct consequence of increased prescription 2007 645 125 (115–135) 101 (93–110) 6.2 (4.1–8.4) 22 (18–26) 67 (60–74) 2008 694 132 (122–142) 107 (98–116) 8.9 (6.4–11.5) 20 (16–24) 72 (65–79) opioid use in the population: a more complex 2009 732 136 (126–146) 110 (101–119) 8.6 (6.1–11.0) 22 (18–26) 75 (68–82) conceptual model, including drug diversion, 2010 722 132 (123–142) 106 (97–114) 10.3 (7.6–12.9) 21 (18–25) 73 (65–80) tampering and other non-recommended 2011 833 150 (140–161) 118 (109–128) 10.5 (7.8–13.2) 29 (24–33) 81 (73–88) use, is required to fully understand how 2012 994 176 (166–187) 143 (133–152) 11.9 (9.0–14.7) 29 (24–33) 98 (90–106) prescribing trends and population-based 2013 1,029 179 (168–190) 144 (134–153) 13.4 (10.4–16.4) 30 (26–35) 105 (96–113) 2014 1,094 187 (176–198) 151 (141–161) 13.5 (10.5–16.5) 31 (27–36) 111 (102–119) exposure affect opioid poisoning trends. The Average % change +6.8% (5.2–8.4%) +6.5% (4.7–8.2%) +11.7% (7.4–16.2%) +8.0% (5.6–10.5%) +8.4% (6.4–10.5%) high proportion of hospital admissions that per year were coded as intentional self-poisoning a: Number of hospital admissions 2006 -14; transfers and in-hospital deaths excluded. suggests that mental health should also b: Rate of hospital admissions related to opioid poisoning, per 1,000,000 person-years 2006 -14. c: Calculated using a Poisson model of admissions per year, adjusted for age and gender; offset by population size. be taken into account. Future research to CI = confidence interval. provide more insight into the trends and potential for prevention and intervention Table 5: Prescription opioid poisoning related deaths in Victoria 2007–2012. could include analysis of coronial data to Prescription opioid poisoning Diagnosis (ICD10) Intentional establish comorbidity, substances involved Other opioids – Methadone Other synthetic – deaths, n=876 self–poisoning, and drug source, and circumstances of the codeine, morphine (T40.3), n=254 narcotics (T40.4), n=143 (T40.2), n=640 n=76 opioid poisoning. Data linkage studies of a b b b b b N Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) Rate (95% CI) drug prescription data, hospital admissions Gender data and death data will further contribute to M 570 23.5 (21.5–25.4) 16.9 (15.3–18.5) 7.0 (6.0–8.1) 1.9 (1.4–2.5) 3.1 (2.4–3.8) knowledge of how drug prescribing trends F 306 12.4 (11.0–13.7) 9.3 (8.1–10.5) 3.4 (2.6–4.1) 1.2 (0.7–1.6) 2.7 (2.1–3.4) affect adverse consequences. Age group 0–24 60 3.8 (2.8–4.7) 2.5 (1.7–3.3) 1.4 (0.8–2.0) – This study’s main limitation relates to 25–44 501 35.1 (32.0–38.1) 24.9 (22.3–27.4) 11.8 (10.0–13.5) 2.4 (1.6–3.2) 3.3 (2.3–4.2) incompleteness in the prescription data. The 45–64 281 23.3 (20.6–26.1) 18.1 (15.7–20.5) 5.2 (3.9–6.4) 2.5 (1.6–3.4) 5.9 (4.5–7.3) following prescriptions are not captured: 65–84 25– 5.2 (3.4–7.0) 4.4 (2.7–6.0) 2.9 (1.5–4.2) 35 ‘private’ prescriptions; prescriptions 85+ <10 dispensed in some hospital settings; Year prescriptions obtained through the 2007 109 21.2 (17.2–25.1) 16.3 (12.8–19.8) 5.4 (3.4–7.4) – 2.9 (1.4–4.4) Repatriation Pharmaceutical Benefits Scheme 2008 151 28.7 (24.1–33.3) 24.2 (20.0–28.4) 7.2 (4.9–9.5) – 3.2 (1.7–4.8) (RPBS) and prescriptions that are below the 2009 154 28.7 (24.1–33.2) 22.7 (18.7–26.7) 6.9 (4.7–9.1) – 6.3 (4.2–8.5) 2010 155 28.4 (23.9–32.9) 19.0 (15.4–22.7) 8.8 (6.3–11.3) 2.7 (1.4–4.1) 4.8 (2.9–6.6) co-payment threshold and paid out of pocket. 2011 158 28.5 (24.1–33.0) 18.8 (15.2–22.4) 10.1 (7.5–12.8) 3.1 (1.6–4.5) 4.5 (2.7–6.3) These omissions may affect the described 2012 149 26.5 (22.2–30.7) 17.6 (14.1–21.0) 8.3 (6.0–10.7) 4.3 (2.6–6.0) 4.6 (2.8–6.4) use per age groups: drugs that are below the a: Number of deaths in 2007-12. co-payment threshold for regular patients b: Deaths per 1,000,000 person-years may appear in the PBS data for concession c: 2007-10 death data are final; 2011 death data has been revised, and the 2012 deaths are provided as preliminary data: these numbers are likely to change when the 2012 revised and final data are released. patients such as pension card holders. In a d: Exact number is not given to prevent recalculation of the low cell count for the ≥85 year age group. study of prescription opioid analgesics in CI = confidence interval. 90 Australian and New Zealand Journal of Public Health 2017 vol . 41 no . 1 © 2016 The Authors Alcohol, Tobacco and other Drugs Prescription opioid dispensing and poisoning 15. Centers for Disease Control and Prevention. Vital signs: Australia, 5.7% of oxycodone dispensings and older Australians; particularly oxycodone Overdoses of prescription opioid pain relievers and in 2008 were reported to be below the co- use among older persons has increased. other drugs among women-United States, 1999-2010. payment threshold, and 2.8% were private Hospital admissions (2006-14) and deaths MMWR Morb Mortal Wkly Rep. 2013;62(26):537-42. 16. Centers for Disease C Control and Prevention. Vital prescriptions. Of morphine dispensings, (2007-12) related to prescription opioid signs: Overdoses of prescription opioid pain relievers- 2.7% were under the co-payment threshold poisoning also increased, but these were United States, 1999-2008. MMWR Morb Mortal Wkly Rep. 2011;60(43):1487-92. and 3.4% were private prescriptions. Tramadol most common in the 25-44 year age group. 17. White AG, Birnbaum HG, Schiller M, Tang J, Katz and codeine had higher proportions of Furthermore, the majority of hospital NP. Analytic models to identify patients at risk for prescription opioid abuse. Am J Manag Care. non-co-payment dispensings: 12.8% and admissions were due to intentional self- 2009;15(12):897-906. 20.1%, respectively, and 4.3% and 6.7% were poisoning. The results emphasise the need 18. Hall AJ, Logan JE, Toblin RL, Kaplan JA, Kraner JC, private prescriptions. In another recent for more in-depth research on opioid-related Bixler D, et al. Patterns of abuse among unintentional pharmaceutical overdose fatalities. J AMA. study, the overall capture of prescription hospital admissions and deaths in Australia, 2008;300(22):2613-20. opioids in the PBS and RPBS in the period and a better understanding of the drivers 19. Hollingworth SA, Gray PD, Hall WD, Najman JM. Opioid analgesic prescribing in Australia: A focus on gender 1992 -11 were reported to be lower (68%); of prescribing and adverse consequences and age. Pharmacoepidemiol Drug Saf. 2015;24(6): 17% were reported to be obtained through of prescribing. Drug diversion, drug 628-36. 20. Blanch B, Pearson SA, Haber PS. An overview of the private prescriptions and 15% were below tampering as well as physical and mental patterns of prescription opioid use, costs and related the co-payment threshold (although these comorbidity need to be taken into account. harms in Australia. Br J Clin Pharmacol. 2014;78(5): are Australia-wide findings that may not In combination with monitoring of opioid 1159-66. 21. Leong M, Murnion B, Haber PS. Examination of opioid accurately reflect the situation in Victoria). prescribing and opioid poisoning, this will prescribing in Australia from 1992 to 2007. Intern Med Furthermore, over-the-counter codeine was facilitate early detection and prevention of a J. 2009;39(10):676-81. 22. Roxburgh A, Burns L, Drummer OH, Pilgrim J, Farrell not captured in this study: in Australia, 55.8% potential opioid epidemic. M, Degenhardt L. Trends in fentanyl prescriptions and of codeine packs sold in 2013 were over- fentanyl-related mortality in Australia. Drug Alcohol Rev. 2013;32(3):269-75. the-counter; for Victoria, this was 55.4%. References 23. Roxburgh A, Bruno R, Larance B, Burns L. Prescription of In other words, the current report under- opioid analgesics and related harms in Australia. Med J 1. Chen LH, Hedegaard H, Warner M. Drug-poisoning estimates actual opioid prescription filling. Aust. 2011;195(5):280-4. deaths involving opioid analgesics: United States, 24. Rintoul AC, Dobbin MD, Drummer OH, Ozanne-Smith 1999-2011. NCHS Data Brief. 2014;166:1-8. This inaccuracy is greater for younger persons J. Increasing deaths involving oxycodone, Victoria, 2. National Center for Health Statistics. Multiple Cause of than for older persons, as older persons Australia, 2000-09. Inj Prev. 2011;17(4):254-9. Death, 1999−2013 Query 2.2.15. In: CDC Wonder Online 25. Roxburgh A, Hall WD, Burns L, Pilgrim J, Saar E, Nielsen Database. Atlanta (GA): Centers for Disease Control and are more likely to receive prescriptions at S, et al. Trends and characteristics of accidental and Prevention; 2015. concessional rates (i.e. for older persons, intentional codeine overdose deaths in Australia. Med 3. National Institute on Drug Abuse. Overdose Death Rates J Aust. 2015;203(7):299. dispensings captured in the co-payment PBS [Internet]. Bethesda (MD): NIDA; 2015 [cited 2015 Sep 26. National Health and Medical Research Council. Chapter 7]. Available from: http://www.drugabuse.gov/related- records better reflect the total numbers of 5.1: Institutional responsibilities. In: National Statement topics/trends-statistics/overdose-death-rates dispensings). Age-differences in prescription on Ethical Conduct in Human Research (2007) (Updated 4. Rudd RA, Aleshire N, Zibbell JE, Gladden RM. Increases May 2015) [Internet]. Canberra (AUST): Government in drug and opioid deaths - United States, 2000-2014. filling are therefore likely to be overestimated of Australia; 2007 [cited 2015 Feb 4]. Available MMWR Morb Mortal Wkly Rep. 2016;64(50);1378-825 in this study. Another limitation is that this from: https://www.nhmrc.gov.au/book/chapter-5-1- 5. Centers for Disease Control and Prevention. Vital Signs: institutional-responsibilities Opioid Painkiller Prescribing. [Internet]. Atlanta (GA): study is based on separate data sources 27. Department of Health and Ageing. Pharmaceutical CDC; 2014 [2015 Sep 10]. Available from: http://www. that were not linked: access to unit level Benefits Scheme (PBS) [Internet]. Canberra (AUST): cdc.gov/vitalsigns/opioid-prescribing/index.html Government of Australia; 2015 [cited 2015 Jan 30]. prescription data and linkage with hospital 6. King NB, Fraser V, Boikos C, Richardson R, Harper S. Available from: http://www.pbs.gov.au/pbs/home Determinants of increased opioid-related mortality in and death data were not possible at the time 28. Mellish L, Karanges EA, Litchfield MJ, Schaffer AL, Blanch the United States and Canada, 1990-2013: A systematic of the study, but this will be the direction of B, Daniels BJ, et al. The Australian Pharmaceutical review. Am J Public Health. 2014;104(8):e32-42. Benefits Scheme data collection: A practical guide for 7. Boudreau D, Von Korff M, Rutter CM, Saunders K, Ray GT, future research in this area. researchers. BMC Res Notes . 2015;8:634. Sullivan MD, et al. Trends in long-term opioid therapy 29. Collaborating Centre for Drug Statistics Methodology. Victoria is the second-most populous state for chronic non-cancer pain. Pharmacoepidemiol Drug International Language for Drug Utilization Research Saf. 2009;18(12):1166-75. of Australia. In a recent study of opioid [Internet]. Geneva (CHE): World Health Organsiation; 8. Caudill-Slosberg MA, Schwartz LM, Woloshin S. Office 2014 [cited 2014 Jun 16]. Available from: http://www. dispensing in Australia, high-to-low ranking visits and analgesic prescriptions for musculoskeletal whocc.no/ pain in US: 1980 vs. 2000. Pain. 2004;109(3):514-9. of opioid dispensed through the PBS/RPBS 30. Australian Bureau of Statistics. Estimated Resident 9. Franklin GM, Mai J, Wickizer T, Turner JA, Fulton-Kehoe per state in 2013 shows Victoria as ranking Population (ERP) by Region, Age & Sex, 2001 to 2013 D, Grant L. Opioid dosing trends and mortality in 32 [Internet]. Canberra (AUST): ABS; 2015; [cited 2015 Washington State workers’ compensation, 1996-2002. fifth out of the eight jurisdictions. Victoria Jan 19]. Available from: http://stat.abs.gov.au/Index. Am J Ind Med. 2005;48(2):91-9. ranked fourth in oxycodone dispensings, aspx?DataSetCode=ABS_ERP_ASGS 10. Johnson H, Paulozzi L, Porucznik C, Mack K, Herter B, 31. Unick GJ, Rosenblum D, Mars S, Ciccarone D. fifth in tramadol dispensings and sixth Hal Johnson C, et al. Decline in drug overdose deaths Intertwined epidemics: national demographic trends after state policy changes - Florida, 2010-2012. MMWR in morphine dispensings. The results in hospitalizations for heroin- and opioid-related Morb Mortal Wkly Rep. 2014;63(26):569-74. overdoses, 1993-2009. PLoS One. 2013;8(2):e54496. presented in this study, although limited to 11. Paone D, Tuazon E, Kattan J, Nolan ML, O’Brien DB, 32. Mofizul Islam M, McRae IS, Mazumdar S, Taplin S, Dowell D, et al. Decrease in rate of opioid analgesic Victoria, do not misrepresent prescription McKetin R. Prescription opioid analgesics for pain overdose deaths - Staten Island, New York City, 2011- opioid dispensing in Australia. management in Australia: Twenty years of dispensing. 2013. MMWR Morb Mortal Wkly Rep. 2015;64(18):491-4. Intern Med J. 2015. doi: 10.1111/imj.12966. 12. Rudd RA, Paulozzi LJ, Bauer MJ, Burleson RW, Carlson 33. Gisev N, Nielsen S, Cama E, Larance B, Bruno R, RE, Dao D, et al. Increases in heroin overdose deaths - Degenhardt L. An ecological study of the extent and 28 States, 2010 to 2012. MMWR Morb Mortal Wkly Rep. Conclusion and implications factors associated with the use of prescription and over- 2014;63(39):849-54. the-counter codeine in Australia. Eur J Clin Pharmacol. 13. Compton WM, Jones CM, Baldwin GT. Relationship Overall prescription opioid consumption 2016;72(4):469-94. between nonmedical prescription-opioid use and heroin use. N Engl J Med. 2016;374(2):154-63. has increased rapidly in Victoria in 2006-13. 14. Rudd RA, Aleshire N, Zibbell JE, Gladden RM. Increases Increase in prescription opioid dispensing in drug and opioid overdose deaths - United States, has been most pronounced among women 2000-2014. MMWR Morb Mortal Wkly Rep. 2016;64(50- 51):1378-82. 2017 vol . 41 no . 1 Australian and New Zealand Journal of Public Health 91 © 2016 The Authors

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Feb 1, 2017

Keywords: ; ;

There are no references for this article.