Quantifying the Adequacy of Opioid Analgesic Consumption Globally: An Updated Method and Early Findings

Abstract

Opioid analgesics are the mainstay for treatment of moderate and severe pain but, in many countries, the consumption of these medicines is inadequate. Over time, various groups have published opioid analgesic metrics, including authors from the World Health Organization. They linked consumption to a level considered adequate based on the actual consumption in developed countries.

In this study, we present our current results on the adequacy of opioid analgesic consumption. We included statistics for 18 controlled opioid medicines that are primarily used as analgesics, and we developed the Adequacy of Opioid Consumption (AOC) Index. The average of the 20 most developed countries for 2015 is set as equal to an AOC Index of 100. An AOC Index of 100 or higher is considered adequate consumption.

The average opioid analgesic consumption of the top-20 countries of the Human Development Index increased from 84 morphine milligram equivalents per capita (2000) to 256 morphine milligram equivalents per capita (2015). The extremes we found for 2015 were Germany (AOC Index: 304) and Nigeria (AOC Index: 0.0069). These extremes differ by 44 000 times. Adequacy of opioid analgesic consumption continues to be problematic around the world.

Opioid analgesics are the mainstay for the treatment of moderate and severe pain. Alternative medicines such as nonsteroidal antiinflammatory medicines are effective in pain management, but have severe gastrointestinal and cardiac adverse effects. In addition, they are often not strong enough to address moderate and severe pain. Some people propose nonpharmacological approaches for chronic pain and even to cut the supply of opioids to chronic pain patients after first putting “supportive measures” in place.¹ However, such supportive measures still require the simultaneous application of an effective therapy and it is not clear what else this should be for most patients, other than opioid analgesics.

Since 1987, the World Health Organization (WHO) has recommended strong opioids for cancer pain and does so to this day.² It also recommends opioid analgesics for pediatric chronic noncancer pain, although it has no guidelines published for acute or chronic noncancer pain in adults.³

For many years, the International Narcotics Control Board (INCB) warned that there is a huge undertreatment of moderate and severe pain around the world.⁴ This was to some degree reflected in opioid use, but very differently across countries. Hence, in 2000, the consumption of opioid analgesics had increased in industrialized countries but not in developing countries.⁵

The main determinant for the level of adequacy of access to opioid analgesics is development level of the country, regardless of whether it is expressed as the Human Development Index (HDI) or gross national product.^6–9 However, although government policies do contribute to development in general terms, it is not clear how policies can contribute to improved access directly. Uganda, for example, had policies aiming at improved access in place for more than 20 years, but it is still at an extremely low level.

For longer times, it has been thought that legislative and regulatory interference in medical practice and patient care is an important barrier for access to opioid analgesics.¹⁰ This may be true for those countries where the only permitted use of opioid analgesics is cancer pain and where, as a consequence, chronic noncancer pain and postsurgical pain are not treated or undertreated. However, for Eastern Europe, Vranken et al. found a low correlation with the number of legal barriers, and the Access to Opioid Medication in Europe Project (ATOME) found that medical education on pain management was an important factor.^11,12 Moreover, cultural and local traditions may also be important impediments to change.

In 2005, the INCB asked WHO “to prepare a technical study on the medical needs for opioids in order to facilitate the identification by Governments of appropriate quantities of opioids required for medical purposes.”^13(p97) In response, Seya et al. developed the Adequacy of Consumption Measure (ACM), a measure indicating adequacy based on opioid consumption statistics in combination with country statistics on morbidity and development level.⁶ Seya et al. published data for 2006.⁶ Later, it was updated by Duthey and Scholten for 2010.⁸

With our current research project, we are further developing the methods for measurement of opioid consumption adequacy, and we analyzed data for 2015 and developments from 2000 to 2015. Monitoring adequacy of opioid consumption is an important first step, allowing us to identify countries where there is insufficient access to adequate pain management. Our work is a continuation of the work by the first author when he was working for WHO.^6,8 Because of the complicated morbidity correction, the ACM does not allow more complicated longitudinal analysis, and, moreover, for some countries, we suspect that health statistics have political bias. Furthermore, limiting the measurement to 5 opioids only might distort the outcome in a negative way for some countries where less-usual opioid analgesics are used. By adapting the method, we assumed that we measure more precisely, and by presenting it as an index with 100 as a standard, we intended to improve comprehensibility.

METHODS

The INCB made its consumption statistics database from 1990 to 2015 (as per July 2017) available for this research project, containing the annual consumption by country and by substance for all substances controlled under the Single Convention on Narcotic Drugs as Amended by the 1972 Protocol, as far as INCB received these data from the countries. All countries are mandated to report the consumption of the substances controlled by the Single Convention to INCB annually, according to their obligations from the convention, which makes these statistics relatively reliable.¹⁴

Substances and Equipotency

The INCB consumption database contains 90 unique substances. As not all substances present in the INCB database are strong opioid analgesics, we excluded those substances from further calculations that (1) were not being used as an opioid analgesic after 1999, (2) were not primarily being used as an analgesic, (3) were generally considered to not be a strong opioid analgesic, and (4) were the active principle of the preparations listed in Schedule III of the Single Convention. (Statistics for such preparations relate to the country where preparations are manufactured and not to where they are consumed; there is no strong opioid analgesic listed in Schedule III.) This resulted in the inclusion of 18 opioid analgesic substances.

Consumption data in the INCB database are presented as a weight. On a weight basis, various substances have different effects. Therefore, to compare the amounts consumed for each substance, they need to be converted to morphine equipotency.¹⁵ For equipotency, we use morphine milligram equivalents (MMEs) as a unit and we based our weight-to-equipotency conversion ratio on defined daily doses (DDDs).

We used DDDs as published by the WHO Collaborating Centre for Drug Statistics Methodology.¹⁶ WHO recommends that, if possible, opioids are not administered parenterally, but orally and, therefore, if a substance has DDDs for more than 1 route of administration, the oral administration route is preferred. If there is no oral DDD, we used the DDD for the transdermal route. The use of MMEs for converting weights into a metric that allows comparison of each study medication was inter alia described by Gilson et al.¹⁵

Twelve of the selected opioid analgesics have an official DDD established by WHO. For 6 of the substances, no official WHO DDD exists.¹⁶ Five of these have an INCB statistical defined daily dosage (S-DDD). We checked these 5 S-DDDs against the literature for plausibility.^17–19 Dependent on the findings, the S-DDD was either used or adapted into a plausible value. For the remaining substance, data on relative potency and dosage were collected from the literature and the product leaflet, and a new DDD was established.^20,21 The selected substances and the DDDs established for this study are represented in Table 1.

TABLE 1—

Opioid Analgesic Substances Selected and Their Defined Daily Doses (DDDs): Quantifying the Adequacy of Opioid Analgesic Consumption Globally

Opioid Analgesic	Official DDD	INCB S-DDD	Other Sources, DDD	DDD Used in This Study
Alphaprodine	. . .	120	240	240
Bezitramide	15	15		15
Dextromoramide	20	20		20
Dihydroetorphine	. . .	. . .	0.2	0.2
Dipipanone	. . .	75	67	75
Fentanyl	1	0.6		1
Hydromorphone	20	20		20
Ketobemidone	50	50		50
Levorphanol	. . .	6	20	20
Morphine	100	100		100
Nicomorphine	30	30		30
Oxycodone	75	75		75
Oxymorphone	. . .	10	13, 33, 67	50
Pethidine	400	400		400
Phenazocine	3	20		3
Piritramide	45	45		45
Tilidine	200	200		200
Trimeperidine	. . .	200		200

Note. INCB = International Narcotics Control Board; S-DDD = statistical defined daily dosage.

Like the ACM as used by Seya et al.⁶ and Duthey and Scholten,⁸ we used for our Index the 20 most developed countries as a comparator. However, our method and the resulting parameter are different, by having 13 more opioid analgesics included and no morbidity correction applied. The new measure was named the Adequacy of Opioid Consumption (AOC) Index.

Data Management

We loaded all data into Stata version 14 (StataCorp LP, College Station, TX) for further analysis. We made some calculations in Office Excel 2017 (Microsoft, Redmond, WA).

Benchmark

For each country in the top-20 HDI, we calculated the sum of the 2015 per-capita consumption of the included opioids, expressed as MMEs. Then we calculated the average 2015 per-capita consumption of these countries. We defined this average as an AOC Index equal to 100, which is the level we used as the benchmark: an AOC Index of 100 or above is considered to be adequate.

The HDI has been published by the United Nations Development Program since 1975, and almost annually since 1990. The HDI is a composite index measuring average achievement in 3 basic dimensions of human development: a long and healthy life, knowledge, and a decent standard of living.²²

Data listed by INCB as “Israel” refer to Palestine and Israel combined; they have an average HDI ranking lower than 20 and data referring to these 2 countries do not therefore qualify for inclusion in the calculation of the benchmark. It was replaced by France, which is next in ranking.

Data Analysis

For all countries, we calculated the sum of the per-capita consumption of the included opioids, expressed as MMEs for the years 2000 to 2015. We obtained the AOC Index for 2015 by dividing the country’s consumption for 2015 by the benchmark as described previously and multiplying it by 100.

We classified adequacy by levels: adequate (AOC Index ≥ 100), moderate (AOC Index < 100 and ≥ 30), low (AOC Index < 30 and ≥ 10), very low (AOC Index < 10 and ≥ 3), and extremely low consumption (AOC Index < 3). With this classification, levels have equal distances on a logarithmic scale.

We used this classification to calculate the number of people living in countries with an AOC Index in each class by WHO region for the year 2015.

RESULTS

The average consumption of the top-20 HDI countries developed from 84 MMEs per capita in 2000 to 256 MMEs in 2015 (Figure 1). The per-capita consumption for 2000 and 2015 is presented in Table 2. Because we defined the average for 2015 as the benchmark in our study (AOC Index of 100), 256 MMEs equals adequacy.

FIGURE 1—

Development of the Average Consumption of Opioid Analgesics in the 20 Most Developed Countries, 2000–2015

TABLE 2—

Percentage of the World Population Living in Countries at the Various Levels of Adequacy of Opioid Analgesic Consumption (AOC): 2015

Consumption Level	AOC Index	%
Adequate consumption	≥ 100	8
Moderate consumption	< 100 and ≥ 30	4
Low consumption	< 30 and ≥ 10	3
Very low consumption	< 10 and ≥ 3	7
Extremely low consumption	< 3	71
No data	. . .	7

For 2015, we found that Germany, Canada, and Austria had the highest adequacy of opioid consumption (AOC Index 304, 271, and 216, respectively), while Nigeria, Myanmar, and Madagascar had the lowest adequacy of opioid consumption (AOC Index 0.0069, 0.0085, and 0.0122, respectively). The difference between the 2 extremes, Germany and Nigeria, is 44 000 times.

Moreover, we calculated that, in 2015, 5.96 billion people (81.6% of the world population) lived in countries with a low consumption level or lower. Only 840 million people (11.5%) lived in countries with a moderate or adequate consumption level. (No data were available for a number of countries with a total population of 500 million [6.8%]; however, most of these countries have characteristics similar to the countries with the lowest consumption levels.) Data are presented in Figure 2.

FIGURE 2—

Global Opioid Analgesic Consumption in (a) 2000 and (b) 2015

Note. HDI = Human Development Index; MME = morphine milligram equivalent. Opioid analgesic consumption increased from 2000 to 2015, but mainly in the most developed countries.

DISCUSSION

Seya et al. based their work on the experience that access to medical and pharmaceutical care is usually best in the most developed countries.⁶ They assumed that this also applies to access to opioid analgesics.

They defined the ACM as an ACM of 1 being equal to the average of the per-capita consumption level for the 20 most-developed countries in 2006. Then, they assumed that an ACM of 1 or higher is adequate. Their method is morbidity-corrected by combining mortality statistics for cancer, HIV, and injuries; the outcomes of studies on the amount of opioids needed per patient for the treatment of terminal cancer, HIV, and for lethal injuries; and 2006 consumption data of opioid analgesics. It measures 5 commonly used opioid analgesics (fentanyl, hydromorphone, morphine, oxycodone, and pethidine). The lowest ACM found by Seya et al. was 0.00014 (unpublished data by Willem Scholten). The discrepancy between the highest and the lowest consumption level was 50 000 times.

Duthey and Scholten updated the ACM for 2010 data and slightly amended the ACM by expressing it as a percentage where 100% and higher represent adequacy.⁸ They found that the most-developed countries had further increased their per-capita consumption and that the discrepancy between the countries with the highest and the lowest ACM had increased to 72 000 times. They also analyzed how adequacy developed from 2006 to 2010: 67 countries at all levels of development increased the adequacy of opioid consumption during this period. The number of people living in countries where adequacy went up from “virtually no consumption” (adequacy < 3%) to “very low consumption” (adequacy of 3%–10%) increased by 236 million people (from 7% to 10%).

Duthey and Scholten concluded that there was some improvement in the adequacy of opioid analgesic consumption around the world, unrelated to development level.⁸ However, their results did not allow them to draw conclusions about how inequality between developing and industrialized countries developed from 2006 to 2010.

By including the 20 most-developed countries in our benchmark, we used the notion that the most-developed countries on average have the best achievements in public health outcomes. However, in many cases, not all are doing equally well, and this is also the case for opioid analgesic consumption. Our benchmark included countries at the highest and at the lowest consumption levels and, therefore, considering an AOC Index of 100 as adequate is not exaggerated (Figure A, available as a supplement to the online version of this article at http://www.ajph.org).

Our findings show that the inequality in adequacy of access to opioid analgesics continues to be huge around the world. The essence remains the same, however. Whereas Duthey and Scholten, using the ACM, found for 2010 that 7.5% of the world population had moderate or adequate access,⁸ we found, using the AOC Index, that this had increased to 11.5 (without a % sign, because of the change in method). However, the increase does not keep pace with the growing world population: in absolute terms, the number of people living in countries with adequacy levels being “low” or lower increased from 5.56 billion in 2010 to 5.96 billion in 2015, an increase of 400 million people who will not receive the pain management they may need when falling ill or needing surgery.

Other author groups analyzed global opioid consumption, too, but did not establish a benchmark for assessing adequacy. The Pain and Policy Studies Group has a longstanding history of research into the reasons for inadequate access to opioid analgesics and examined trends and factors influencing the consumption.^5,7,10,15 Usually, it calculates opioid metrics for 5 commonly used strong opioids.

As mentioned in the introduction, the work of Seya et al.⁶ was a response to a request by the INCB to WHO to define the medical needs for opioids. However, INCB did not use these WHO findings when they became available, but established its own benchmark when it conducted a longitudinal study on opioid analgesic consumption from 2001 to 2013. Over 2010 and 2015, INCB published reports stating that consumption is inadequate below a level of 200 S-DDDs per million inhabitants per day.^22,23 By doing so, INCB also implied that the consumption is adequate above this level. However, the INCB report provides no justification for this benchmark. It should also be noted that 200 S-DDD per million inhabitants per day is much lower than an ACM of 100% (adequate level) and corresponds with an ACM₂₀₁₀ of 3.4%, which is classified by Seya et al.⁶ and by Duthey and Scholten⁸ as “very low consumption.” In a 2016 journal publication in parallel to INCB’s 2015 report, however, authors from INCB called this level “low,” but not inadequate; but again, no justification was provided.⁹

The INCB longitudinal study found that the opioid analgesic use more than doubled worldwide between 2001 to 2003 and 2011 to 2013, but with countries in some regions having no substantial increases. Gross domestic product and HDI were the main determinants for intercountry differences.

Our AOC Index is a quick indicator, useful for policy purposes, showing the magnitude of adequacy. It is not an exact indication of country needs and it does not allow one to evaluate adequacy of pain management at the individual or hospital level. For the latter purpose, monitoring pain levels in individual patients is more appropriate. Furthermore, the method does not allow for identifying overconsumption in a country. Although we have no indications that in any country overall per-capita consumption is too high, it is possible that in some countries, some people with pain are prescribed too-high quantities, while others are withheld adequate pain management.

In our study, we included all controlled opioids used around the world as analgesics primarily. The INCB included 12 opioids in its longitudinal study. Seya et al.,⁶ Duthey and Scholten,⁸ and the Pain and Policy Study Group⁵ included only the 5 main opioid analgesics (which constitute about 80% of all opioid analgesic consumption). It requires further analysis whether this results in significant differences. We do not expect that they will affect most countries, but some countries may have different patterns of use. If in these countries larger quantities of other opioid analgesics are used (e.g., ketobemidone), this could change the picture for these countries, and if these countries are part of the top-20 most developed countries, they are influencing the benchmark as well.

The INCB statistics do not show the indication for which opioids are used and include only substances regulated under the Single Convention on Narcotic Drugs. The mandate of INCB is the result of negotiations between countries 60 years ago and there is no easy way for INCB to adapt data collection to current needs. Such needs may include making a distinction between the type of use: medical indication, scientific use, or even diversion from pharmacies. Hence, our data source is somewhat coarse and, combined with our exclusion criteria, it has a consequence that opioids such as codeine (meeting exclusion criteria 3 and 4), tramadol and buprenorphine (not in the INCB database), and opioids mainly used for other purposes such as anesthesia (remifentanil, sufentanil) or treatment of opioid dependence (heroin, methadone) are not included in our analysis. We have to accept that worldwide statistics provide limited information. Those who make an analysis on the country or hospital level could make such distinctions more easily when collecting their data.

With our study, we identified countries and regions where pain management is inadequate and where opioid analgesic consumption should increase. It is not meaningful to warn countries with such low consumption levels against a similar epidemic of opioid-induced deaths as currently taking place in the United States, like other authors did.²⁴ On the contrary, even in the United States, the epidemic is mainly driven by illicitly produced opioids without any medical purpose, as the CDC has admitted.²⁵ Indeed, already in 2010, there was evidence that the epidemic was not mainly driven by regularly prescribed opioid analgesics.^26,27 However, the US situation is often incorrectly interpreted. Hence, policy measures are taken that do not adequately counteract the causes of the epidemic, but instead affect adequate pain management.

With our study, we have shown that the adequacy of opioid analgesic consumption continues to be problematic around the world. Also in 2015, the large majority of people (almost 6 billion) live in countries where access is inadequate, meaning that they will not have their pain managed when falling ill. This number increased over the years, despite an increase in the number having moderate or adequate access. It shows that improvement of adequacy is limited to a number of countries and that policy efforts do not manage to keep pace with the increasing world population.