Addressing Problematic Opioid Use in OECD Countries

3. The opioid crisis and the rise of an epidemic in some OECD countries

3.1. What is the magnitude of the opioid crisis in OECD countries?

Several different countries have had epidemics associated with the use of opioids. For instance, France, Portugal and Switzerland had a sizeable heroin epidemic in the 1980s and 1990s (EMCDDA, 2000[16]); and the United States had opiate epidemics in the late 19th century, after World War II and in the 1960s (Lawson, 2018[17]). Today’s situation is challenging because it involves both prescribed and illicit opioids at the same time. The predominance of use and harms of one versus the other varies across OECD countries, but the threat now also comes from the global diffusion of information and commercial exchanges.

In terms of prescription analgesic opioids, at the global level there is a lack of access to pain relief and palliative care, including a staggering lack of access to opioids in low income countries (Knaul et al., 2018[18]). In OECD countries, the reality is quite different. The average availability, defined as the amounts that each country’s competent national authority estimates are used annually (including reporting of medicine destroyed, losses during manufacturing, etc.), has been steadily growing in the past 15 years (see Figure 3.1). The sharpest increases happened in the 2000s, where between 2002‑04 and 2005‑07 analgesic opioids availability grew on average by 58.6%. More recently, between 2011‑13 and 2014‑16, the growth rate dropped to 5.4% on average. It is important to highlight that this data does not directly reflects the consumption of analgesic opioids in countries, but the general availability for different purposes, which the largest component is for medical use.

In the triennium 2014‑16, among the countries above the OECD average in availability of analgesic opioids, only the United States (-12.9%), Belgium (-7.3%), Denmark (-18.2%) and Australia (-10.9%) have reduced use. The countries who experienced an increase of over 50% between 2011‑13 and 2014‑16 are Israel (125%), the United Kingdom (67.8%), Slovakia (64.9%), Greece (53.9%), Portugal (56.3%) and Colombia (76.6%), but the latter four countries remain below the OECD average.

The increasing availability of analgesic opioids in OECD seems to be associated with increasing medical prescription of opioids in health systems. Other than the United States and Canada, different studies have shown increasing trends of medical opioid prescription in the past 10 to 15 years in countries such as France (Chenaf et al., 2019[19]); Germany (Schubert, Ihle and Sabatowski, 2013[20]); Italy (Musazzi et al., 2018[21]); Netherlands (Wagemaakers et al., 2017[22]); Poland (Dzierżanowski and Ciałkowska-Rysz, 2017[23]); Spain (AEMPS, 2019[24]); United Kingdom (Zin, Chen and Knaggs, 2014[25]); Australia (AIHW, 2018[26]); Denmark, Norway and Sweden (Muller et al., 2019[27]). It is relevant to note that although prescriptions may be rising, in some countries this might not necessarily equate to a rise in dosages as they can be for lower doses and/or quantities of opioids.

Figure 3.1. Mean availability of analgesic opioids in OECD countries 2011‑13 and 2014‑16. S-DDDs per million inhabitants per day

Note: Analgesic opioids include codeine, dextropropoxyphene, dihydrocodeine, fentanyl, hydrocodone, hydromorphone, morphine, ketobemidone, oxycodone, pethidine, tilidine and trimeperidine. It does NOT include illicit opioids.

S-DDD: Defined daily doses for statistical purposes.

Source: (INCB, 2018[28]).

Statlink

Capturing the size of the illicit opioids market is very difficult. Seizures of drugs are one way of approaching it, but important limitations should be noted when interpreting this data. For instance, seizures depend on law enforcement, customs regulations and capacity across countries. Countries can be producers or transit countries, not necessarily consumers of the seized opioids. Between 2012 and 2016, opioids seizures have been relatively stable in OECD countries – particularly in the last three years (see Figure 3.2) – with 2013 as the year with the highest record of over 80 000 kilograms seized in total and 2012 with the lowest level, with almost 47 000 kilograms.

Figure 3.2. Trend in kilograms of opioids seized in OECD countries 2012‑16

Note: The blue line refers to the right vertical axis and the bars to the left vertical axis.

Source: (UNODC, 2018[29]).

Statlink

Taking the average of seizures per million inhabitants between 2012 and 2016 (see Figure 3.3), Turkey, Mexico, Hungary, Greece, Korea, United States, Australia, Netherlands and Belgium are above the OECD average. Comparing the periods of 2013‑14 and 2015‑16, Mexico, Korea, France, Luxembourg, Colombia, Austria and Canada have increased their opioids seizures, according to this indicator, and Turkey, Hungary, Greece, United States, Australia, Netherlands and Belgium have decreased.

Figure 3.3. Average annual kilograms of opioids seized per million inhabitants in OECD countries 2012‑16

Source: (INCB, 2018[28]).

Statlink

Opioid-related deaths (ORD) is a key indicator that reflects the impact of problematic use on population health and, at the same time, how health systems and other related government services are performing in this area. In 25 OECD countries for which data is available, the average of ORD has increased by 20% in recent years (see Figure 3.4). Among the countries above the average, the United States, Canada, Sweden, Norway, Ireland, and England & Wales have seen increasing trends.

Figure 3.4. Opioid-related deaths per million inhabitants, 25 OECD countries, 2011‑16

Note: Countries ranked by latest year with available information.

Source: EMCDDA for European countries and country responses to ORD data questionnaire.

Statlink

Men represent the largest share of ORD in Europe, accounting for three out of four deaths in the last five years for which information is available (see Table 3.1).

Table 3.1. Gender distribution of opioid-related deaths in European countries with data, 2012‑16

	2012	2013	2014	2015	2016
Number of countries	18	18	19	18	14
Male deaths	1 494	3 095	3 445	3 612	1 219
Female deaths	412	1 022	1 117	1 218	297
Total deaths	1 906	4 117	4 562	4 830	1 516
M/F ratio	3.63	3.03	3.08	2.97	4.10
% males	78.38%	75.18%	75.52%	74.78%	80.41%

Note: 2012 and 2016 does not include the UK, which represents 36‑40% of total deaths in the other years.

Source: EMCDDA.

The opioid crisis has unfolded in different magnitudes across OECD countries. For instance, in Australia there were 1 119 opioid-induced deaths. More than 900 of these included mention of prescription opioids. After adjusting for age, there was a 62% rise between 2007 and 2016 (from 2.9 to 4.7 deaths per 100 000 people) but it is still lower than the peak in 1999. The rate in men was also 2.1 times as high as for women (AIHW, 2018[26]). The majority of these deaths (76%) were attributable to prescription opioids. Figure 3.5 compares the data on prescription opioids between Australia and Canada in 2017, showing that codeine ranks first in both countries, hydromorphone plays a larger role in Canada, and that tramadol and buprenorphine are more common in Australia (CIHI, 2018[30]).

Figure 3.5. Number of defined daily doses dispensed of prescription opioids Australia and Canada, 2017

Note: DDD – daily defined dose.

Source: (CIHI, 2018[30]).

In Canada, there were 10 337 ORD between January 2016 and September 2018, with death rates increasing from 8.4 per 100 000 population in 2016 to 11.1 in 2017 and 11.8 in 2018. Canada registered 4 034 apparent opioid-related deaths in 2017, 3 017 in 2016 and 3.286 between January and September 2018. Among these deaths, around 73‑74% involved fentanyl or its analogues, and the great majority were accidental deaths (Special Advisory Committee on the Epidemic of Opioid Overdoses, 2019[31]). In relation with years-of-life-lost, life expectancy at birth was analysed in British Columbia, showing that life expectancy decreased by 0.38 years from 2014 to 2016, and fatal drug overdoses (the majority involving opioids) accounted for 32% of the decrease (Ye et al., 2018[32]).

In the United States, 399 230 people have died from an opioid overdose between 1999 and 2017, while in 2015‑17 life expectancy fell for the first time in more than 60 years largely as a result of the opioid crisis. Moreover, prescription rates appear to be higher where labour force participation is lower, showing that the dislocation of opportunities is also associated with the opioid crisis (OECD, 2018[33]). The cost of the opioid drug epidemic in 2015 was USD 504 billion or 2.8% of GDP that year, showing the great economic impact of the crisis (Council of Economic Advisers, 2017[34]).

Figure 3.6. Overdose deaths involving opioids, by type of opioid, United States, 1999‑2016

Source: (CDC-NCHS, 2018[35]).

Statlink

A particularly relevant group of the population are pregnant women and new-borns. Opioid use among pregnant women is associated with the neonatal abstinence syndrome (NAS), which refers to a postnatal opioid withdrawal syndrome that can occur in 55 to 94% of new-borns (McQueen and Murphy-Oikonen, 2016[36]). In the United States, the use of opioids among pregnant women has been rising, affecting more women who are non-Hispanic White, with less educational and income level, and who have health insurance coverage. In addition, pregnant opioid-polydrug users are likely to report past-year anxiety/depression and are also most likely to report past alcohol/drug use treatment, reflecting the important role played by mental health in the crisis (Metz et al., 2018[37]).

3.2. Factors underpinning the development of the opioid crisis

The current opioid crisis has been caused by the co-occurrence of several factors over the years, both from the supply and demand sides. The main determinants of the crisis can be summarised into four factors, as follows.

3.2.1. Increased opioids prescription and over-prescription in health systems

The use of opioids is useful for pain management, particularly for acute pain, which can be derived from many sources including poor or unstable material conditions of life and mental health issues, not just physical symptoms. Pain was a neglected issue before the 1990s, with health care professionals focussing more on whether they can cure or treat the underlying problem, rather than manage symptoms that affect the well-being of the person involved. The factors that caused this neglect have, however, changed over time.

Uncorroborated claims about the safety and risks of prescription opioids

In 1980, a letter published in a prestigious journal (Porter and Jick, 1980[38]) concluded that “despite widespread use of narcotic drugs in hospitals, the development of addiction is rare in medical patients with no history of addiction”. A bibliometric analysis (Leung et al., 2017[39]) revealed that it was ‘heavily and uncritically cited as evidence that addiction was rare with long-term opioid therapy’. A second study (Portenoy and Foley, 1986[40]), which analysed a sample of only 38 patients, described opioid maintenance therapy as a safer and ‘more humane’ alternative for patients with intractable non-terminal pain and no history of problematic drug use. Despite these reports, which were widely accepted by the medical community, high-quality evidence of opioids’ safety in terms of addictive effects did not exist at the time (Christie et al., 2017[41]). Given the lack of evidence about adverse effects, in the 1990s pain was called the ‘fifth vital sign’ encouraging health care professionals to assess pain more widely and urged more aggressive use of opioids for chronic non-cancer pain. This contributed to create a culture that was more prone to the use of opioids (Kolodny et al., 2015[42]).

Opioid manufacturers and advocacy groups have influenced pain management

Opioid manufacturers have played a significant role in escalating opioid prescription. In the United States, manufacturers have funded pain advocacy organisations, medical societies, clinical practice guideline development efforts, and medical education (US Senate, 2017[43]). They have also developed marketing campaigns spreading the message that opioids were low-risk medications and effective at managing a wide range of chronic pain conditions (Van Zee, 2009[44]). Subsequently, advocacy groups have petitioned for the prescription of opioids in several ways, including issuing guidelines recommending the use of opioids for pain management and opposing efforts to monitor and regulate opioid over prescription (Whyte, Mulvihill and Wieder, 2016[45]) (see Box 3.1).

Box 3.1. Influence of prescription opioid manufacturers in the United States

The influence of pharmaceutical manufacturers on pain management advocacy groups and prescribers has been considerable. During the late 1990s and the 2000s, opioids manufacturers conducted marketing campaigns, targeted mainly at physicians and patients, downplaying the problematic effect of opioids arguing that concerns over dependence and other dangers from the drugs were overstated (Van Zee, 2009[44]).

Opioid manufacturers have directed considerable amounts of financial resources to different actors involved in the market of prescription drugs. A report from the United States Senate Committee on Homeland Security and Governmental Affairs found that opioid manufacturers contributed USD 9 million to 14 third-party advocacy organisations between 2012 and 2017, and have destined USD 1.6 million in payment to physicians affiliated with these advocacy groups (McCaskill, 2018[46]). The report signals that initiatives from the advocacy groups often echoed and amplified messages favourable to increased opioid use. Another study found that physicians who received any opioid-related payments from industry had 9.3% more opioid claims compared to physicians who received no such payments, and that each additional industry-sponsored meal received was associated with an increase of 0.7% in opioid claims (Hadland et al., 2018[47]). Along this line, a study reported that between 1 August, 2013, and 31 December, 2015, there were USD 39.7 million in non-research-based opioid marketing distributed to 67 507 physicians across 2 208 US counties and found that increased county-level opioid marketing was associated with elevated overdose mortality one year later, an association mediated by opioid prescribing rates (Hadland et al., 2018[47]).

Recently, there has been a widespread negative reaction to the 2016 CDC Guideline for Prescribing Opioids for Chronic Pain. One study (Lin et al., 2017[48]) shed light on the correlation between the opposition to the guideline and the financial relationships with opioid manufacturers, concluding that ‘opposition to the guideline was significantly more common among organisations with funding from opioid manufacturers than those without funding from the life sciences industry’.

In 2007, one of the main opioid manufacturers pled guilty in federal court to overstating the benefits and understating the dependence risk of an extended release formulation of oxycodone. The settlement of USD 600 million is one of the largest in history with a drug company (US Western District Court of Virginia, 2007[49]). Similarly, in March 2019 the same manufacturer reached a USD 270 million settlement with the state of Oklahoma to avoid going to a state court trial over the company’s role in the opioid crisis. Despite the several government and class action settlements against opioid companies in the past 15 years, opioid litigation has not come near the USD 13 billion-a-year opioid industry (Haffajee and Mello, 2017[50]).

Poor opioid prescribing practices and insufficient education in pain management

Opioids overprescribing is considered one of the most important root causes of the crisis. In the United States alone, there were 240 million opioid prescriptions dispensed in 2015, nearly one for every adult in the general population (Makary, Overton and Wang, 2017[51]). Among overprescribers, three group of doctors can be identified: physicians who appear to have a doctor–patient relationship with people who deceive them into treating them as patients in pain (e.g. “doctor shopping strategies”); physicians who treat patients in pain but with high doses of opioids; and doctors who abuse their privileges and knowingly arrange for opioids to be taken by people who don’t necessarily need them (e.g. ‘pill mills’) (Reidenberg and Willis, 2007[52]). In addition, health care purchasers have generally failed to influence or regulate the prescription and use of opioids. In the United States, for example, a recent study found that many insurers failed to apply "utilisation management" rules (e.g. prior authorisation, quantity limits, and cost sharing) to discourage opioid overuse and encourage safer and more effective alternatives (Lin et al., 2018[53]). Another study highlighted the policy adopted by some payers (including Medicaid) that encouraged doctors to prescribe methadone due to its low cost as a risk factor for overprescribing (Webster et al., 2011[54]).

Medical education in pain management, opioid prescribing and screening for substance use disorders has been inadequate in institutions and medical schools (Chiu et al., 2018[55]; Webster et al., 2017[56]). While high-quality evidence on the direct relationship between insufficient physician education and the opioid crisis is still lacking, inadequate education has been hypothesised to influence opioid overprescribing (Christie et al., 2017[41]).

Limited alternatives for chronic pain management and lack of insurance coverage

Alternative treatments for pain management, such as non-steroidal anti-inflammatory drugs (NSAIDs), gabapentoids, antidepressants and muscle relaxants, are not always effective. Opioid analgesics are often used when these other treatments have not worked (Kroenke and Cheville, 2017[57]). Unfortunately, research & development on non-addictive treatments for chronic pain has not received the same priority as other areas (Charumilind et al., 2018[58]). In addition, non-pharmacological interventions such as exercise, multidisciplinary rehabilitation, acupuncture, cognitive behavioural therapy, and mind-body practices have been associated with durable slight to moderate improvements in function and pain for specific chronic pain conditions, but these are less utilised or lacking insurance coverage (Skelly et al., 2018[59])

Defective approach to mental health and opioid use

Mental health conditions and their treatments can interact with opioid use, but these interrelations are not always appropriately addressed. For instance, in the United States, 18.7% of all patients with mental health conditions receive 51.4% of the total opioid prescriptions distributed each year, meaning that having a mental health disorder was associated with a two-fold greater use of prescription opioids (Davis et al., 2017[13]). Risk for opioid overdose is higher for individuals with depression and prescribed opioids, while longer duration of benzodiazepine use is associated with drug overdose. Notably, antidepressant use for more than 90 days is associated with reduced odds of overdose for persons with depression (Bair and Bohnert, 2015[60]).

3.2.2. A dynamic illicit drugs market has fuelled the crisis

The illicit trade in drugs has been growing globally, boosting supply. Opioids, including the clandestine development of synthetic chemicals, like fentanyl and its analogues, are a significant driver of this growth.

High purity and increasing availability of illicit opioids at a low cost

The current opioid crisis has also been linked with the increased availability and purity of opiates (e.g. heroin) at lower prices. Total global opium production in 2017 was the highest estimate recorded by UNODC since it started estimating global opium production. Opium prices fell in Afghanistan (the largest producer in the world) by 47% from December 2016 to December 2017 and the price of high-quality Afghan heroin decreased by 7% over the same period (World Drug Report, 2018[8]). In the United States, heroin purity has increased by almost 40% from the 1980s to 2000, while its price has decreased, going from over USD 3 200 per gram in 1981, to just over USD 600 per gram in 2012 (US Drug Enforcement Administration, 2017[61]). Clandestinely produced fentanyl and fentanyl analogues are newcomers in the opioids landscape, where, for instance, carfentanil seizures had a tenfold increase and fentanyl a fourfold increase in 2016 (World Drug Report, 2018[8]).

Small shipments constitute a relatively new source for trafficking drugs, where online sales and shipment via the postal service or express consignment appear to be driving to some extent the increases (OECD, 2018[62]). Figure 3.7 shows that narcotics top the list of the most frequent type of seizures among 15 OECD countries.

Figure 3.7. Most frequent types of seizures of small shipments

Note: The figures on the left scale correspond to the number of countries that mentioned the most frequent types of seized illicit goods.

Source: Based on fifteen country responses to OECD Survey (OECD, 2018[62]).

Statlink

The darknet is as a relatively new market for accessing opioids and drugs (Box 3.2). According to a recent survey to 15 OECD countries, the most commonly seized products via the darknet were narcotics, confirming the use of small shipments to spread risk across a wider range of shipments and to avoid large-scale seizures of illicit narcotics (Figure 3.8).

Figure 3.8. Most common forms of illicit trade via “darknet”

Note: The left scale corresponds to the number of countries that mentioned the types of illicit goods most commonly seized involving the “darknet”.

Source: Based on fifteen country responses to OECD Survey (OECD, 2018[62]).

Statlink

Box 3.2. The “darknet” as a small but growing source for illicit drugs, including opioids

Darknet is a term associated with the use of online platforms that anonymise the identities of users and vendors. It shares many characteristics with legal online marketplaces, including the provision of digital open markets in which geographically dissimilar vendors advertise goods and customers provide scores to rank sellers´ product quality and service. Untraceable crypto-currencies are often used to avoid financial scrutiny, which is why they are also called “cryptomarkets” (Finklea, 2017[63]).

A darknet study conducted by European agencies (EMCDDA and Europol, 2017[64]) found that more than 60% of all listings on five major darknet markets worldwide up to August 2017 were related to the illicit sale of drugs. They also estimated that world drug sales on the darknet from November 2011 to February 2015 amounted to roughly EUR 44 million per year. However, in early 2016 drug sales were between EUR 150 million and EUR 270 million per year (Kruithof et al., 2016[65]). Regarding the darknet market for opioids, US-based vendors comprise 36% of global opioid transactions, followed by the UK (16%), France (14%), Germany (12%), the Netherlands (9%) and Australia (9%) (Martin et al., 2018[66]).

However, the sale of online drugs remains comparatively small. The global darknet drug market is estimated to account for no more than 0.1–0.2% of the combined annual drug retail markets of the US and the EU (World Drug Report, 2018[8]). The online Global Drug Survey, based on a non-representative convenience sample of around 100 000 self‑selected people in over 50 countries, found that the proportion of Internet users who purchased their drugs via the darknet rose from 4.7% in 2014 to 9.3% in 2018. The highest proportions of Internet users reporting the purchase of drugs via the darknet were found in North America, Oceania and Europe (Winstock et al., 2018[67]).

Polysubstance use and problematic use

Polysubstance use combining different types of opioids with benzodiazepines, alcohol (e.g. binge drinking), psychoactive prescriptions (perhaps due to mental health issues), and “top up” opiate in addition to prescriptions are associated with fatal overdoses (Frisher et al., 2012[12]). Early polysubstance use (e.g. youth) often sets the stage for a later transition from medical to problematic use of opioids prescribed for pain. However, polysubstance use is a problem across the lifespan. For example, in 2010, the percentage of emergency department visits in the U.S. that involved prescription opioids and alcohol was highest among persons aged 30‑44 years (20.6%) and 45–54 years (20.0%). Among deaths due to prescription opioids, persons aged 40–49 years (25.2%) and 50–59 years (25.3%) had the highest percentage of alcohol involvement (Jones, Paulozzi and Mack, 2014[68]).

Prison and jail post-release period

Prisons and jails are a common factor in opioid related deaths and overdoses. In fact, prisons present a striking concentration of people with OUD. While the prevalence rate of OUD in Europe is less than 1% among the general public, the rate reaches 30% in the prison population. Given that heroin-related fatal overdoses occur mainly after a period of abstinence, opioid users are particularly at risk during the post-release period. The highest risk of fatal overdose episodes is registered during the first two weeks after release, where the risk of overdose death increases more than sevenfold (Strang et al., 2016[69]).

3.2.3. Poor treatment and actions to minimise the negative consequences for OUD patients

Several interventions have proven to be effective to treat OUD. However, there are barriers to their use, both in health systems and in other sectors (e.g. criminal justice). This lack of access to treatment has contributed to the crisis.

Barriers to access medication assisted therapy

There is strong evidence showing that medication assisted therapy (MAT) significantly reduces the risk of mortality. Compared with patients receiving MAT, untreated patients have 2.2 to 3.2 times higher risk of all-cause mortality and 4.8 to 8.1 times higher risk of overdose mortality. Retention in MAT of over 1-year is associated with lower mortality rate than that with retention of less than one year, meaning that long-term care is beneficial for patients (Ma et al., 2018[70]; Sordo et al., 2017[71]). Barriers for MAT access can be classified into financial/economic aspects and governmental support (e.g. insufficient funding, lack of non-economic governmental support, economic crisis); formularies (e.g. reimbursement issues, insufficient service provision, shortage of palliative care experts, inadequate integration of MAT into primary care); education and training (e.g. insufficient under-, post-graduate and continuing education); and societal attitudes (e.g. fear, lack of awareness, inadequate information, stigma) (Larjow et al., 2016[72]; Kolodny and Frieden, 2017[73]; Maksabedian Hernandez, 2017[74]).

Predominance of abstinence-only rehabilitation therapies

In the United States and Canada, rehabilitation programmes are still mainly abstinence based (Annan et al., 2017[75]). More specifically, in the United States only 8‑9% of all substance treatment facilities between 2006 and 2016 had MAT programmes certified by SAMHSA. The proportion of all clients receiving methadone ranged from 23‑30% and between 1‑5% for buprenorphine in the same period (Substance Abuse and Mental Health Services Administration, 2017[76]). This happens despite evidence showing that opioid users who are treated only with psychological support are at twice greater overdose death risk than those who receive opioid agonist pharmacotherapy (Pierce et al., 2016[77]).

Inadequate access to evidence-based harm minimisation interventions

Harm minimisation interventions such as needle and syringe programmes (Abdul-Quader et al., 2013[78]) and naloxone availability for overdose management (Chimbar and Moleta, 2018[79]) have substantive evidence supporting their effectiveness. Similar is the situation for supervised drug consumption rooms (EMCDDA, 2018[80]) , where some countries have also implemented them. Despite this, access to these interventions has been lacking or could be improved in many countries. For instance, inadequate responses by people who witness overdose episodes and lack of access to naloxone has contributed to the increase in opioid related deaths. This aspect is critical, since more than half of all fatal overdoses occur in the victim’s home and more than half of deaths occur with another person present. It is estimated that one in four fatalities could have been prevented if the witness had acted differently (Frisher et al., 2012[12]; Strang et al., 2016[69]).

3.2.4. Social and economic conditions contributing to the crisis

The environment in which people live is linked to drug use, including the consumption of opioids. Social and economic conditions, particularly of vulnerable groups of the population, have contributed to the crisis.

Unemployment appears to be linked to the opioids issue

A recent study in the United States found that as county unemployment rates increase by one percentage point, the opioid death rate per 100 000 rises by 3.6% and the opioid overdose emergency department visit rate per 100 000 increases by 7% (Hollingsworth, Ruhm and Simon, 2017[81]). The relationship between problematic opioid use and unemployment, however, is very complex and some studies suggest a reverse causality, claiming that it is problematic use that leads to an increase in unemployment. For instance, Krueger (2017[82]) indicates that the increase in opioid prescriptions between 1999 and 2015 could have accounted for 20% of the decline in the prime-age male and 25% of the prime-age female labour force participation rate over those years. Another study (Gitis and Soto, 2018[83]) came to similar conclusion linking a decline in labour force participation in the United States due to OUD.

Economic recession has also been correlated with an increase in teenage illicit drug use (Arkes, 2007[84]) and in self-reported substance-use disorders related to analgesics (including opioid and non-opioid forms) (Carpenter, McClellan and Rees, 2017[85]). Economic recessions and unemployment influence illegal drug use through three mechanisms (Nagelhout et al., 2017[86]). First, people may use drugs to cope with the psychological distress caused by unemployment; second, drug use could be motivated by the increase in time available; and third, drug use could be motivated by the social exclusion incurred as a consequence of the loss of social status caused by unemployment.

Lack of housing affecting the most vulnerable population

The evidence on the relationship between drug use and housing condition has explored primarily the influence of the former on the latter, identifying drug use as one of the causes of homelessness and unstable housing (Zerger, 2012[87]). Nonetheless, the literature available on the impact of housing conditions on drug use highlights that the opposite causal relationship also holds true, with an unstable housing situation increasing problematic use of opioids and other drugs. Unstable housing prevents people who use drugs from accessing treatment, both by discouraging people to engage with health services (Burkey, A. Kim and Breakey, 2011[88]) and by posing practical obstacles, such as lack of medical insurance. Drug and alcohol use is sometimes a necessary social ritual to be accepted among the homeless community (Zerger, 2012[87]). Unstable housing negatively affects treatment retention rates and effectiveness. Finally, it also exacerbates psychiatric symptoms, which are often a co-occurring condition affecting many drug users (Fox et al., 2016[89]).

Social stigma as a relevant barrier for prevention and recovery

The role of stigma towards people who use drugs is considered as a barrier to individual care/treatment, as well as a barrier to societal support for broader policy shifts. Research indicates that stigma contributes to individuals poor mental and physical health, non‐completion of substance use treatment, delayed recovery and reintegration processes, and increased involvement in risky behaviour (e.g. needle sharing) (Livingston et al., 2012[90]). Stigma themes include individual perceptions of opioid dependence, community perceptions of opioid dependence, blame as a stigmatising factor, language surrounding opioid use, and treatment experience (Cooper and Nielsen, 2017[91]).

Box 3.3. Exploring the factors associated with opioid-related deaths in OECD countries

There have been few studies exploring the magnitude of the opioid crisis at the international level, and much less revising the potential factors related to its impact on population health across countries. In this context, an assessment of whether a relationship exists between a range of societal factors and opioid-related deaths (ORD) rates in 25 OECD countries was conducted.

Methods included using national level data for variables such as household income and savings rates, unemployment, poverty, GDP per capita, divorce rate, health and social spending, incarceration rates and country governance indicators. A series of econometric techniques were applied in order to fit the best models that could identify the factors associated to ORD. Finally, fixed regression analysis was performed with the most suitable variables identified through econometric criteria and literature review.

The findings were not sufficiently robust to draw meaningful scientific conclusions nor policy considerations. In particular, the differences of ORD data collection across countries limits the analysis, so relevant efforts should be placed on improving the information infrastructure to better capture drug-related deaths and other harms data, including incidence of problematic opioid users, health services utilisation, recovery rates, etc.