country need for opioids

Journal of Pain & Palliative Care Pharmacotherapy. 2011;25:6–18.Copyright © 2011 Informa Healthcare USA, Inc.ISSN: 1536-0288 print / 1536-0539 onlineDOI: 10.3109/15360288.2010.536307

ARTICLE

A First Comparison Between the Consumption of and theNeed for Opioid Analgesics at Country, Regional, andGlobal Levels

Marie-Josephine Seya, Susanne F. A. M. Gelders, Obianuju Uzoma Achara,Barbara Milani, and Willem Karel Scholten

ABSTRACT

The objective of this study was to propose a rough but simple method for estimating the total population need foropioids for treating all various types of moderate and severe pain at the country, regional, and global levels. Wedetermined per capita need of strong opioids for pain related to three important pain causes for 188 countries.These needs were extrapolated to the needs for all the various types of pain by using an adequacy level derivedfrom the top 20 countries in the Human Development Index. By comparing with the actual consumption levelsfor relevant strong opioid analgesics, we were able to estimate the level of adequacy of opioid consumptionfor each country. Good access to pain management is rather the exception than the rule: 5.5 billion people(83% of the world’s population) live in countries with low to nonexistent access, 250 million (4%) have moderateaccess, and only 460 million people (7%) have adequate access. Insufficient data are available for 430 million(7%). The consumption of opioid analgesics is inadequate to provide sufficient pain relief around the world. Onlythe populations of some industrialized countries have good access. Policies should seek a balance betweenmaximizing access for medical use and minimizing abuse and dependence. Countries should aim to increasethe medical consumption to the magnitude needed to address the totality of moderate and severe pain.

KEYWORDS Access, accidents and injuries, cancer, consumption, demographics, education and training,health policy, HIV/AIDS treatment, legislation (health), need, opioid, pharmaceuticals products, policy, statis-tics, substance abuse

Dr. Marie-Josephine Seya and Dr. Obianuju Uzoma Achara are affiliatedwith School of Pharmacy, Temple University, Philadelphia, Pennsylvania,USA. Dr. Marie-Josephine Seya is also affiliated with the BMS VirologyPolicy and Advocacy, Rutgers University Ernest Mario School of Phar-macy, Plainsboro, New Jersey, USA.

Dr. Susanne F. A. M. Gelders is an independent consultant in Maas-tricht, The Netherlands.

Dr. Barbara Milani and Dr. Willem Karel Scholten are affiliated withEssential Medicines and Pharmaceutical Policies, World Health Organiza-tion, Geneva, Switzerland.

This study was done as part of WHO’s Access to Controlled MedicationsProgramme. It was funded through grants from the US Cancer Pain Re-lief Committee and the World Health Organization, which for this purposewas also supported by the Dutch Ministry of Health. The WHO Collab-orating Centre for Policy and Communication in Cancer Care providedopioid consumption data. The authors would also like to thank Dr ColinMathers for his advice on the selection and collection of statistical data,Dr Ahmadreza Hosseinpoor for his assistance with statistical calculations,Dr Richard Laing and Professor Albert I. Wertheimer for commenting onthe manuscript, and Dr Philip Jenkins for editing a previous version of themanuscript.

Address correspondence to Willem Karel Scholten, PharmD, MPA, EssentialMedicines and Pharmaceutical Policies, World Health Organization, Geneva,1211 Switzerland (E-mail: [email protected]).

INTRODUCTION

Oral opioids are key components for the treatment ofmoderate to severe pain, and several are regarded asessential medicines (1). Despite the effectiveness ofopioid analgesics, many people live in pain becausethey do not receive appropriate treatment. The re-alization of Millennium Development Goal (MDG)8e, i.e., “. . . to provide access to affordable essen-tial drugs in developing countries,” is likely to be fur-ther away for opioid analgesics than for any otherclass of medicines. Pain can vary from mild to severe,and different pain levels require different analgesics.Paracetamol, acetylsalicylic acid, nonsteroidal anti-inflammatory medicines (NSAIMs) and opioids witha ceiling effect (often called “weak-acting opioids”)are not effective for moderate to severe pain. NSAIMscan have serious side effects, leading to concerns over

6

Journal of Pain & Palliative Care Pharmacotherapy 7

the long-term safety of chronic use. Despite a centuryof medical chemistry, suitable alternatives to strongopioids for treatment of moderate to severe pain havenot been found.

Indeed, several barriers contribute to the insuffi-cient utilization of opioid analgesics. These barriersare multifaceted and may be related to legal aspects,policy, knowledge, or attitudes. In many countriesseveral of these aspects are present. Most of thesebarriers trace back to fears of abuse and dependenceof opioids, particularly the fear of prescribed opi-oids being diverted to illicit circuits (2, 3). Therefore,many laws and government policies primarily focuson rendering opioids unavailable, without acknowl-edging that their rational medical use is beneficial topatients in pain (2, 4, 5). Still today, in many coun-tries, patient access to opioids for cancer pain is pro-foundly restricted through legislation (6). However,for terminal patients dependence is irrelevant. More-over, during treatment of chronic noncancer pain,dependence rarely results (7). Diversion of prescrip-tion medicines from domestic distribution channelshas been reported (8). However, this is not a rea-son not to prescribe opioids to patients. Reports fromIndia and Malaysia show that diversion by patientsis rare or nonexistent (9, 10). When patients can-not access pain relief, this may result in excruciatingsuffering, suicide, or the use of street-bought heroin(11).

In recent years the problem of inadequate pain re-lief has attracted more and more the attention of theinternational community. United Nations (UN) bod-ies, including the World Health Assembly, the Eco-nomic, Social and Cultural Council, the Commissionon Narcotic Drugs, and the International NarcoticsControl Board (INCB), declared that access shouldimprove (9, 12–15). Since 2006, the INCB has re-quested annually that all governments promote ratio-nal medical treatment with narcotic drugs and psy-chotropic substances (8, 16–19). There is recogni-tion now that pain relief is part of the human right tothe highest attainable standard of mental and physicalhealth, or is even a human right on its own (20–22).Some countries (e.g., Uganda, Romania, Colombia)have made serious efforts to improve opioid accessi-bility (23–26).

A comparison of consumption of and need foropioids is an important tool to recognize inadequa-cies of access to opioids as medicine. Various simplemethods have been suggested to establish the mor-phine needs of a country (27–29). However, thesemethods fail to take into account the specific mor-bidity patterns in the countries themselves and donot relate to the adequacy of the actual use of opioidanalgesics.

METHODS

At country level, we collected mortality data forcancer, human immunodeficiency virus (HIV), andinjuries, and calculated the per capita amounts ofopioids that would be sufficient to treat the pain re-lated to these diseases (“Need of morphine equiva-lents (selected diseases) in mg per capita”). Cancerand HIV/AIDS (acquired immunodeficiency syn-drome) are the two major causes of chronic severemalignant pain, and injuries are an important causeof acute moderate to severe pain. Pain can also havemany other causes. We also calculated the actual percapita consumption of relevant strong opioids (mor-phine, fentanyl, oxycodone, hydromorphone, andpethidine) from the per capita consumption of eachof them (“Consumption of morphine equivalents inmg per capita”). These statistics are relatively reliable,because governments cannot import strong opioidswithout submitting these figures to the INCB.

No objective standard exists for an adequate levelof opioid consumption, and we therefore had to de-velop such a standard ourselves. The selection ofreference countries should be independent from theopioid consumption. Therefore, we took the top20 countries from the Human Development Index(HDI) for 2004 as reference countries (30). We ar-bitrarily took their average ratio between calculatedper capita consumption and per capita need as a stan-dard for an adequate consumption level. We nor-malized the ratio by designating this average to bean “adequacy of consumption measure” (ACM) of1.00. Furthermore, we calculated for each countrythe amount of opioids that could be adequate for paintreatment (“Adequate consumption (all pain condi-tions) in kg”), and from this we calculated the globalrequirements.

Data Collection

We collected data from 188 countries. For the ac-tual consumption figures, we used 2006 data from theINCB for all relevant strong opioids and calculatedtheir joint consumption in “morphine equivalents,”i.e. the dosage of a substance that equals the analgesicpotency of 1 mg oral morphine (using multiplicationfactors inversely proportional to their Defined DailyDosages). We excluded methadone, because this isusually used for treatment of opioid dependence.

For each country, we retrieved population data(2006), the age-standardized mortality rate for can-cer (2002), and the age-standardized mortality ratefor lethal injuries (2002) from the World HealthOrganization (WHO) Statistical Information System(WHOSIS) Database. For deaths due to HIV/AIDS

C© 2011 Informa Healthcare USA, Inc.

TA

BL

E1a

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Afr

ican

Reg

ion

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Alg

eria

5.95

0.44

4530

0.00

33L

esot

ho50

.82

No

data

2316

No

data

Ang

ola

16.1

40.

0261

040.

0000

Lib

eria

∗11

.45

No

data

936

No

data

Ben

in12

.23

0.14

2447

0.00

05M

adag

asca

r8.

540.

0537

370.

0003

Bot

swan

a41

.16

4.15

1747

0.00

44M

alaw

i28

.58

0.01

8858

0.00

00B

urki

naF

aso

11.9

00.

0239

040.

0001

Mal

i11

.78

0.02

3220

0.00

01B

urun

di13

.86

0.12

2587

0.00

04M

auri

tani

a10

.30

0.29

716

0.00

13C

amer

oon

17.6

8N

oda

ta73

41N

oda

taM

auri

tius

4.63

6.13

132

0.05

80C

ape

Ver

de∗∗

6.88

0.67

820.

0042

Moz

ambi

que

30.5

90.

3114

,65

40.

0004

Cen

tral

Afr

ican

Rep

ublic

28.4

5N

oda

ta27

71N

oda

taN

amib

ia36

.19

1.57

1692

0.00

19

Cha

d12

.31

No

data

2944

No

data

Nig

er11

.04

0.04

3464

0.00

02C

omor

os8.

65N

oda

ta16

2N

oda

taN

iger

ia14

.19

No

data

46,89

9N

oda

taC

ongo

16.6

00.

0913

990.

0002

Rw

anda

16.0

00.

0434

590.

0001

Cot

ed’

Ivoi

re20

.82

0.06

8996

0.00

01S

aoT

ome

and

Pri

ncip

e∗∗7.

230.

0726

0.00

04

Dem

.Rep

.of

the

Con

go14

.11

0.02

19,54

70.

0000

Sen

egal

9.47

No

data

2612

No

data

Equ

ator

ial

Gui

nea

15.2

0N

oda

ta17

2N

oda

taS

eych

elle

s∗∗7.

116.

0314

0.03

71

Eri

trea

11.5

20.

1712

350.

0007

Sie

rra

Leo

ne12

.72

No

data

1668

No

data

Eth

iopi

a∗10

.79

0.02

19,96

20.

0001

Sou

thA

fric

a31

.16

5.39

34,36

70.

0076

Gab

on20

.06

No

data

601

No

data

Sw

azila

nd8.

79N

oda

ta22

8N

oda

taG

ambi

a10

.74

No

data

408

No

data

Tog

o13

.00

0.18

1903

0.00

06G

hana

11.9

33.

8862

680.

0142

Uga

nda

18.6

40.

7912

,72

60.

0019

Gui

nea

11.0

7N

oda

ta23

22N

oda

taU

nite

dR

ep.o

fT

anza

nia

20.5

4N

oda

ta18

,50

9N

oda

ta

Gui

nea-

Bis

sau

14.4

0N

oda

ta54

1N

oda

taZ

ambi

a34

.97

0.82

9342

0.00

10K

enya

21.3

62.

3417

,83

50.

0048

Zim

babw

e53

.36

No

data

16,12

0N

oda

taW

HO

AF

RO

regi

on17

.06

1.03

301,

500

0.00

17

∗ Cal

cula

tion

sba

sed

onH

IVda

tafo

r20

07fr

omth

eW

HO

Sta

tist

ical

Info

rmat

ion

Sys

tem

(WH

OS

IS)

Dat

abas

e.∗∗

Cal

cula

tion

sba

sed

onca

ncer

and

inju

rym

orta

lity

only

(no

data

for

HIV

mor

talit

yav

aila

ble)

.∗∗

∗ Mor

talit

yda

tafo

rS

erbi

aw

ere

quot

edfr

omJa

ncov

icet

al.(

32)

(Tab

le1d

).∗∗

∗∗C

alcu

lati

ons

base

don

HIV

data

from

UN

AID

SC

ount

ryfa

ctsh

eet

2008

for

Indi

a(T

able

1e).

∗∗∗∗

∗ In

Aus

tria

mor

phin

eis

also

used

for

opio

idag

onis

tth

erap

y(T

able

1d).

8

TA

BL

E1b

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Am

eric

anR

egio

n

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Ant

igua

and

Bar

buda

∗∗7.

80N

oda

ta15

No

data

Guy

ana

10.1

05.

2117

00.

0226

Arg

enti

na8.

072.

2672

120.

0122

Hai

ti12

.41

0.15

2678

0.00

05B

aham

as12

.84

28.7

796

0.09

81H

ondu

ras

9.26

No

data

1475

No

data

Bar

bado

s14

.06

No

data

94N

oda

taJa

mai

ca9.

816.

2860

50.

0280

Bel

ize

14.7

3N

oda

ta95

No

data

Mex

ico

4.99

0.76

11,9

990.

0067

Bol

ivia

14.2

1N

oda

ta30

35N

oda

taN

icar

agua

6.86

1.13

867

0.00

72B

razi

l7.

9810

.83

34,5

220.

0594

Pan

ama

7.55

3.18

567

0.01

85C

anad

a7.

8145

6.59

5810

2.56

01P

arag

uay

7.98

0.35

1097

0.00

19C

hile

7.76

6.51

2920

0.03

67P

eru

10.1

6N

oda

ta64

05N

oda

taC

olom

bia

7.00

3.09

7289

0.01

93S

aint

Kit

tsan

dN

evis

∗∗5.

86N

oda

ta7

No

data

Cos

taR

ica

7.12

4.34

715

0.02

67S

aint

Luc

ia∗∗

7.00

8.01

260.

0501

Cub

a7.

333.

5618

870.

0212

Sai

ntV

ince

ntan

dth

eG

rena

dine

s∗∗

8.40

1.36

230.

0071

Dom

inic

a∗∗7.

805.

4312

0.03

05S

urin

ame

13.9

81.

1814

50.

0037

Dom

inic

anR

epub

lic9.

640.

9221

170.

0042

Tri

nida

dan

dT

obag

o11

.49

No

data

348

No

data

Ecu

ador

7.42

1.24

2238

0.00

73U

nite

dS

tate

sof

Am

eric

a7.

4342

0.70

51,4

012.

4787

ElS

alva

dor

6.78

2.87

1047

0.01

85U

rugu

ay10

.90

No

data

829

No

data

Gre

nada

∗∗10

.77

5.72

260.

0233

Ven

ezue

la6.

601.

0741

020.

0071

Gua

tem

ala

5.79

0.94

1722

0.00

71W

HO

AM

RO

regi

on7.

5117

1.40

153.

600.

9437

See

Tab

le1a

for

defin

itio

nsof

foot

note

s.

9

TA

BL

E1c

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Eas

tern

Med

iter

rane

anR

egio

n

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Afg

hani

stan

8.67

No

data

5169

No

data

Om

an∗∗

5.69

4.10

331

0.03

15B

ahra

in∗∗

6.88

9.11

116

0.05

80P

akis

tan

5.90

0.07

21,69

20.

0005

Djib

outi

11.4

1N

oda

ta21

3N

oda

taQ

atar

∗∗4.

076.

6176

0.07

10E

gypt

4.89

0.85

8289

0.00

76S

audi

Ara

bia∗∗

5.93

5.65

3272

0.04

18Ir

an(I

slam

icR

epub

licof

)6.

242.

0910

,02

20.

0146

Som

alia

9.54

No

data

1840

No

data

Iraq

∗∗6.

13N

oda

ta39

89N

oda

taS

udan

9.31

0.27

8020

0.00

13Jo

rdan

∗∗7.

8310

.39

1025

0.05

81S

yria

nA

rab

Rep

ublic

∗∗3.

272.

8414

480.

0381

Kuw

ait∗∗

4.23

4.94

269

0.05

12T

unis

ia4.

593.

6310

710.

0346

Leb

anon

5.25

5.71

486

0.04

76U

nite

dA

rab

Em

irat

es∗∗

5.44

2.29

528

0.01

84

Lib

yan

Ara

bJa

mah

iriy

a∗∗4.

304.

7659

30.

0485

Yem

en∗∗

5.89

0.17

2923

0.00

12

Mor

occo

3.78

0.47

2664

0.00

54W

HO

EM

RO

regi

on6.

001.

3374

.04

0.00

86

See

Tab

le1a

for

defin

itio

nsof

foot

note

s.

10

TA

BL

E1d

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Eur

opea

nR

egio

n

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Alb

ania

∗∗8.

352.

5460

50.

0133

Lat

via

10.1

921

.31

533

0.09

16A

ndor

ra∗∗

6.82

42.9

712

0.27

58L

ithu

ania

9.11

24.6

270

90.

1184

Arm

enia

9.59

0.98

660

0.00

45L

uxem

bour

g8.

9510

5.36

940.

5153

Aus

tria

∗∗∗∗

∗7.

2232

8.23

1373

1.99

13M

aced

onia

(the

form

erY

ugos

lav

Rep

ublic

of)

8.21

1.27

382

0.00

68

Aze

rbai

jan

6.46

0.23

1239

0.00

15M

alta

8.40

11.7

378

0.06

12B

elar

us∗

8.16

1.09

1815

0.00

59M

oldo

va7.

431.

6765

10.

0098

Bel

gium

8.35

219.

7919

901.

1518

Mon

aco∗∗

6.50

No

data

5N

oda

taB

osni

aan

dH

erze

govi

na∗∗

6.56

No

data

588

No

data

Net

herl

ands

8.72

101.

3232

620.

5087

Bul

gari

a∗6.

8212

.20

1198

0.07

84N

orw

ay7.

7615

5.92

827

0.88

02C

roat

ia∗∗

9.05

32.7

394

10.

1584

Pol

and

10.0

942

.80

8787

0.18

58C

ypru

s∗∗5.

0913

.89

980.

1194

Por

tuga

l7.

9265

.84

1913

0.36

42C

zech

Rep

ublic

9.92

55.7

423

090.

2459

Rom

ania

∗7.

800.

0838

370.

0005

Den

mar

k9.

3829

0.79

1163

1.35

76R

ussi

anF

eder

atio

n∗9.

151.

3429

,94

70.

0064

Est

onia

∗9.

4424

.63

289

0.11

42S

anM

arin

o∗∗7.

57N

oda

ta5

No

data

Fin

land

6.58

174.

0779

11.

1580

Ser

bia/

Mon

tene

gro∗∗

∗7.

2712

.81

1635

0.07

72F

ranc

e7.

7614

5.15

10,87

40.

8187

Slo

vaki

a∗∗9.

2112

0.98

1133

0.57

52G

eorg

ia6.

622.

3567

00.

0155

Slo

veni

a9.

0110

4.09

412

0.50

58G

erm

any

7.97

381.

6615

,03

92.

0972

Spa

in7.

2613

1.53

7277

0.79

32G

reec

e7.

4992

.49

1902

0.54

10S

wed

en6.

6216

1.47

1372

1.06

82H

unga

ry∗

10.9

369

.14

2510

0.27

71S

wit

zerl

and

6.62

216.

7611

271.

4337

Icel

and

9.05

126.

8862

0.61

38T

ajik

ista

n5.

23N

oda

ta79

4N

oda

taIr

elan

d8.

5110

0.21

821

0.51

55T

urke

y∗∗5.

156.

7687

010.

0574

Isra

el∗

7.30

64.3

111

350.

3858

Tur

kmen

ista

n∗∗5.

390.

1260

30.

0009

Ital

y7.

4233

.94

9963

0.20

02U

krai

ne9.

173.

0397

490.

0145

Kaz

akhs

tan

9.45

No

data

3304

No

data

Uni

ted

Kin

gdom

8.07

99.1

411

,15

90.

5376

Kyr

gyzs

tan

6.11

0.28

734

0.00

20U

zbek

ista

n4.

360.

1426

880.

0014

WH

OE

UR

Ore

gion

7.89

87.1

915

9,76

30.

4699

See

Tab

le1a

for

defin

itio

nsof

foot

note

s.

11

TA

BL

E1e

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Sou

thea

stA

sian

Reg

ion

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Ban

glad

esh

6.39

0.99

22,76

00.

0068

Mya

nmar

8.73

0.05

9650

0.00

03B

huta

n6.

45N

oda

ta96

No

data

Nep

al7.

070.

2844

660.

0017

Dem

ocra

tic

Peo

ple’

sR

epub

licof

Kor

ea∗∗

5.54

0.78

3002

0.00

62S

riL

anka

6.76

1.67

2964

0.01

09

Indi

a∗∗∗∗

7.34

0.13

193,

184

0.00

08T

haila

nd8.

122.

4611

,76

80.

0133

Indo

nesi

a7.

240.

2937

,86

90.

0018

Tim

or-L

este

∗∗6.

44N

oda

ta16

4N

oda

taM

aldi

ves∗

7.81

No

data

53N

oda

taW

HO

SE

AR

Ore

gion

7.28

0.34

285,

975

0.00

20

See

Tab

le1a

for

defin

itio

nsof

foot

note

s.

12

TA

BL

E1f

.A

dequ

acy

ofC

onsu

mpt

ion

Mea

sure

(AC

M)

ofC

ount

ries

inth

eW

HO

Wes

tern

Pac

ific

Reg

ion

Cou

ntry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)C

ount

ry

Nee

dof

mor

phin

eeq

uiva

lent

s(s

elec

ted

dise

ases

)in

mg

per

capi

ta

Con

sum

ptio

nof

mor

phin

eeq

uiva

lent

sin

mg

per

capi

ta(2

006)

Ade

quat

eco

nsum

ptio

n(a

llpa

inco

ndit

ions

)in

kgA

CM

(200

6)

Aus

tral

ia7.

2214

2.68

3383

0.86

58N

ewZ

eala

nd∗

7.61

72.1

871

90.

4154

Bru

nei

Dar

ussa

lam

7.86

3.96

690.

0221

Pal

au∗∗

4.99

3.45

20.

0302

Cam

bodi

a11

.88

No

data

3852

No

data

Pap

uaN

ewG

uine

a8.

32N

oda

ta11

79N

oda

ta

Chi

na8.

102.

0424

5,89

20.

0110

Phi

lippi

nes

5.28

0.43

10,41

10.

0035

Fiji

6.35

No

data

121

No

data

Rep

ublic

ofK

orea

9.50

15.6

310

,42

70.

0720

Japa

n6.

4820

.22

18,94

20.

1366

Sam

oa∗∗

5.15

1.98

220.

0168

Kir

ibat

i∗∗2.

82N

oda

ta6

No

data

Sin

gapo

re7.

268.

7272

70.

0526

Lao

Peo

ple’

sD

emoc

rati

cR

epub

lic

8.52

0.29

1120

0.00

15S

olom

onIs

land

s∗∗4.

88N

oda

ta54

No

data

Mal

aysi

a8.

075.

5448

160.

0300

Ton

ga∗∗

4.61

2.84

110.

0270

Mar

shal

lIs

land

s∗∗6.

781.

869

0.01

20V

anua

tu∗∗

4.99

0.67

250.

0059

Mic

rone

sia

(Fed

erat

edS

tate

sof

)∗∗

5.04

3.08

130.

0267

Vie

tN

am7.

190.

8114

,15

40.

0049

Mon

golia

16.9

21.

0110

060.

0026

WH

OW

PR

Ore

gion

7.87

5.50

316,

961

0.03

02

See

Tab

le1a

for

defin

itio

nsof

foot

note

s.

13

14 M.-J. Seya et al.

we used data for 2007 from the UNAIDS Reporton the Global AIDS Epidemic 2008, except wherethe WHOSIS Database contained these data onHIV/AIDS. For India we used data from the UN-AIDS Country Factsheet 2008 (figures for 2007) (seefootnotes, Tables 1a–1f). Mortality data for Serbiawere quoted from Jancovic et al. (2006) (31). WhereAIDS mortality data are presented as “less than,” weused this limit for our calculations (e.g., if mortalitywas “≤10,” we used 10 for our calculations).

Calculation Methods

We calculated the need for morphine required forpain relief in terminal cancer patients, HIV patients,and lethal injuries patients in mg per capita, basedon the assumptions that 80% of terminally ill can-cer patients and 50% of AIDS patients will require75 mg of morphine per person daily for an averageof 90 days during the last year of their lives and that15% of patients with lethal injuries will require 75 mgof morphine daily for an average of 5 days at the endof their lives (32).

This method of calculating morphine need doesnot account for pain from many other causes, includ-ing nonlethal cancers, nonlethal injuries, non–end-stage HIV, surgery, sickle cell episodes, childbirth,chronic nonmalignant pain, and many more. The ac-tual extent of pain from these other diseases is difficultto quantify because of a lack of comprehensive epi-demiological data at the national level and of data onhow much morphine per patient is required on aver-age for these conditions. Therefore, we had to correctfor these other pain causes. Our correction is based onthe assumption that total morbidity in each countryis proportional to the morbidity for the three causesmentioned above. The correction factor is the aver-age of the ratios of (actual consumption:calculatedneed) for the highest 20 countries in the HDI. Theaverage ratio was 22.84. A consumption level equalto or higher than this average is assumed to be an ad-equate consumption level and we “normalized” theaforementioned ratio by introducing an ACM pro-portional to the ratio, designating the average ratio of22.84 to be an ACM of 1.00. Consequently, an ACMof 1.00 or more represents a consumption level re-lated to adequate access to opioid analgesics. Further-more, we defined an ACM ≥0.30 and <1.00 as mod-erate consumption, ≥0.10 and <0.30 as low, ≥0.03and <0.10 as very low, and <0.03 as virtually nonex-istent. We calculated the ACM for all countries andthe relation between the HDI and the logarithm ofthe ACM (log(ACM)).

We also calculated the required absolute consump-tion in order to achieve an ACM of 1.00 at the coun-

try level, for each WHO region and for the world,based on the actual consumption and each region’sACM. For countries with an ACM higher than 1.00,we took the actual consumption as adequate.

RESULTS

For each country, in Tables 1a to 1f we present theper capita need of morphine equivalents and the ac-tual per capita opioid consumption, together with ad-equate consumption in kg and the ACM. Our calcu-lation method is based on a number of assumptions(for instance, that the top 20 HDI countries have anopioid analgesics consumption that is more or less ad-equate to their need) and it does not take into ac-count that variations in morbidity patterns betweencountries may lead to different needs for opioid anal-gesics. Therefore, the results should be consideredan indication of the magnitude of adequacy and notan exact indication of the country needs. Thus, theyshould not be used to calculate the health care sys-tem’s requirements for opioids, as our method doesnot take into account the capacity of the health caresystem. However, we recommend that the ACM andthe calculated adequate consumption will be used forpolicy purposes and in setting mid- and long-termtargets.

We were able to calculate the ACM for 145 coun-tries. The countries for which the ACM could notbe calculated did not report their consumption datato the INCB. We fear that many of these countriesalso have very minimal consumption, if any. The dif-ferences between countries are so large that the re-sults best can be expressed on a logarithmic scale: thecountry with the highest ACM in our study (Canada)has a per capita consumption 50,000 times higherthan the lowest country (Malawi).

Figure 1 shows the relation between the log (ACM)and the HDI for 139 countries. There is a close rela-tionship between the development of a country andthe log (ACM). Most people who live in countrieswhere they have adequate access live in the more in-dustrialized regions and, conversely, worst access isfound in developing countries. However, some indus-trialized countries also have inadequate consumption.

Table 2 presents the number of people living in theWHO regions based on their ACM. From this tablewe conclude that:

5.5 billion people (83% of the world’s population) livein countries with low to nonexistent access;

250 million (4%) have moderate access;460 million (7%) have adequate access;for 430 million (7%) insufficient data are available.

Journal of Pain & Palliative Care Pharmacotherapy


-5.00

-4.00

-3.00

-2.00

-1.00

0.00

1.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

HDI

log

(A

CM

)

FIGURE 1. Relation between the log (ACM) and the Human Development Index (HDI) for139 countries. (Function formula: log (ACM) = −6.4113 + 6.200 × HDI; N = 139; correlationcoefficient: .895; P value: <.0001.)

In 2006, the world used 231 tonnes of morphineequivalents. If all countries increased their consump-tion to adequate levels, the required amount wouldbe 1292 tonnes, or almost 6 times higher.

DISCUSSION

The method we developed and applied for all relevantstrong opioids and to all countries, also taking intoaccount morbidity levels, allows insight into whichcountries need to improve pain treatment and by whatmagnitude. So far there has been no overview avail-able of the total per capita consumption for all com-bined strong opioids. Furthermore, the lack of accessto opioid analgesics for pain treatment was known tobe considerable, but up till now the scale of the prob-

lem was not known to the extent that we have demon-strated in this study. We are aware that improvementof our method is possible, but because no similarmethod is available, we wanted to start a discussionand encourage that others build on our method.

Study Limitations

Our method relies on an extrapolation based on theassumption that total pain prevalence is proportionalto the morbidity from three diseases that resultin moderate to severe pain. For the three diseasesthat we selected, most of the suffering is relatedto the end-of-life stage and, therefore, our choiceis reasonable. It should be borne in mind that incountries where HIV or cancer mortality is extremelyhigh, the method will exaggerate the opioid need.

TABLE 2. Number of People (in Thousands) Living in Countries, According to Adequacy of Consumption Measure (ACM) andRegion

ACM

AFROpopulation(thousands)

AMROpopulation(thousands)

EMROpopulation(thousands)

EUROpopulation(thousands)

SEAROpopulation(thousands)

WPROpopulation(thousands)

Global populationin thousands (%)

ACM ≥ 1 (adequateconsumption)

0 335418 0 128622 0 0 464040(7%)

0.3 ≤ ACM < 1 (moderateconsumption)

0 0 0 227658 0 24670 252328(4%)

0.1 ≤ ACM < 0.3 (lowconsumption)

0 0 0 127390 0 127953 255343(4%)

.0.03 ≤ ACM <0.1 (very lowconsumption)

1338 206346 76506 94160 0 78566 456916(7%)

ACM < 0.03 (virtually noconsumption)

502501 303900 399919 283081 1718985 1510365 4718751(72%)

No data 269953 49280 63858 25944 2063 21810 432908(7%)Total 773792 894944 540283 886855 1721048 1763364 6580286(100%)



However, owing to underreporting, morbidity statis-tics usually underestimate the actual morbidity, inparticular in developing countries. Thus, for suchcountries, the needs we calculated will be too low.Also, we excluded methadone, because it is mostlyused for treatment of opioid dependence. Therefore,our method underestimates the adequacy of opioidanalgesic consumption for the few countries thatuse methadone mainly for pain management Severalopioids are also used in anesthesia. A validation ofour method could be part of future work.

We related our standard of adequacy to the top20 countries of the HDI. We realize that taking thetop 10 or the top 30 would have influenced the out-come considerably. Not taking the latter is supportedby the finding that the countries for which more im-pediments are reported rank 25 and higher, and fromthe countries ranking 1 to 20, hardly any is reportedas having high access barriers (6). The former (i.e.,taking the top 10) might be considered, but in thatcase we would be setting the targets for improvementeven higher for those countries with lower levels ofadequacy of opioid consumption.

Relation to Other Medicines’ Accessibilityand MDG 8e

Only 460 million people live in countries with ad-equate consumption levels. This means that the re-maining 6 billion will not be treated adequately whenin pain. Cameron et al. collected availability datafrom 36 countries for a standard list of 15 core genericmedicines and found a public sector availability rang-ing from 9.7% (Yemen) to 79.2% (Mongolia). Al-though the level of availability cannot be compareddirectly with the consumption level, merely consider-ing that 72% of the world’s population live in coun-tries that do not consume strong opioids, it is notrash to conclude that the situation for strong opioidanalgesics is much worse than it is for these 15 coremedicines in general (33). Therefore, we concludethat MDG 8e on access to essential medicines is fur-ther away for opioid analgesics than for other classesof essential medicines.

Explanation for Inadequacy of OpioidAnalgesic Consumption

Patients have a right to be treated with controlledmedicines listed in the WHO Model List of EssentialMedicines and a right to be protected against drugabuse and dependence (2, 20). Therefore, drugcontrol and public health policies should seek theoptimum public health outcome, which is a balancebetween maximizing access for legitimate medical

use and minimizing abuse, dependence, and diver-sion. However, this is not currently the case in mostcountries. Exaggerated fear that pain patients willbecome dependent and that prescribed opioids willbe diverted from their intended use prevents patientsfrom receiving any pain treatment at all. There is nodoubt that this balance should be ensured.

National policies, laws, and regulations often havea number of drawbacks that hamper adequacy of painmanagement. Lack of medical knowledge and bias to-wards pain management with opioids also affect theadequacy of opioid consumption. There is no eco-nomic reason why countries should not go through atransition towards adequate management of pain, asthe prices in developed countries for morphine tabletsand methadone syrup at supplier level are only a fewUS cents per unit. However, studies and surveys havereported that opioid analgesic cost at the patient levelis higher in developing countries than in developedcountries and this makes them unaffordable outsidehealth system treatment programmes. The cost ofopioid analgesics can be a limiting factor in devel-oping countries where palliative care programs andpain relief are not subsidized by national health sys-tems and where the market size is based on the out-of-pocket purchasing power of patients (34, 35).

Working on Improvement

We suggest that countries that wish to improve ac-cess to these medicines take our figures as a work-ing hypothesis to measure adequacy of access. For allcountries with low access, this will work, because thedifference between their current level and adequateconsumption is so large. At the country level, we rec-ommend that governments take the outcome of ourmethod, which represents approximately the magni-tude needed to address the totality of moderate andsevere pain, as a long-term target for the developmentof the estimates that they have to submit to the INCBannually and strive for an adequate consumption levelof opioid analgesics. This means that they will reach alevel related to an ACM of 1.00 or more. For the esti-mates themselves, the INCB and WHO are currentlypreparing a Joint INCB-WHO Manual for CalculatingEstimates for Drugs Under International Control and werecommend that countries apply this as soon as it isready. Governments should not submit much higherestimates than the country can absorb for rational useby the health system, in order to prevent diversion toillicit circuits. However, estimates should always beslightly higher than the amount that actually will berequired, in order to ensure that importations will notbe blocked.



Consumption will not increase spontaneously, butonly if countries address actively all barriers thatimpede access to adequate pain management. Gov-ernments can do so by implementing the guidelineson balanced controlled substances policies that theWHO Access to Controlled Medications Programme(ACMP) will publish approximately simultaneouslywith this article. These guidelines will address all rel-evant policy aspects of controlled medicines access.The guidelines will include a practical checklist andwill also be available for interested individuals.

WHO has developed the ACMP in consulta-tion with the INCB.34 WHO will manage theACMP, which will support countries when improv-ing access to opioid analgesics and other medicinescontrolled under the international drug control con-ventions. Among other activities, the ACMP is devel-oping treatment and policy guidelines and can assistwhen reviewing policies and legislation. INCB rec-ommends that countries work with the ACMP andprovide it with sufficient resources to reach its objec-tives (8).

Declaration of interest

The authors report no conflicts of interest. The au-thors alone are responsible for the content and writ-ing of this paper.

REFERENCES

[1] World Health Organization. WHO Model List of EssentialMedicines, 15th List. Geneva: World Health Organization;2007. Available at: http://www.who.int/medicines/publications/08 ENGLISH indexFINAL EML15.pdf. Accessed August2009.

[2] Ajaji IO, Soyannwo OA, Amanor-Abadu SD. Availability of opi-oids for cancer pain in Nigeria. Nigerian J Surg. 2000;7:25–28.

[3] International Narcotics Control Board. Availability of Opiatesfor Medical Needs. Report for 1995 (Special report preparedpursuant to Economic and Social Council resolutions 1990/31and 1991/43). New York: International Narcotics ControlBoard, United Nations; 1996. Available at: http://www.incb.org/incb/en/annual report 1995.html. Accessed 2 December 2010.

[4] Blenghini C, Joranson DE, Ryan KM. Italy reforms nationalpolicy for cancer pain relief and opioids. Eur J Cancer Care.2003;12:28–34.

[5] Davis MP, Walsh D. Epidemiology of cancer pain and fac-tors influencing poor pain control. Am J Hospice Palliat Med.2004;21:137–142.

[6] Cherny NI, Baselga J, DeConno F, Radbruch L. Formularyavailability and regulatory barriers to accessibility of opioid forcancer pain in Europe: a report from the ESMO/EAPC OpioidPolicy Initiative. Ann Oncol. 2010;21:615–626.

[7] Noble M, Tregear SJ, Treadwell JR, Schoelles K. Long-termopioid therapy for chronic non-cancer pain: a systematic reviewand meta-analysis of efficacy and safety. J Pain Symptom Man-age. 2008;35:214–228.

[8] International Narcotics Control Board. Report for 2009. NewYork: United Nations; 2010. E/INCB/2008/1.

[9] Rajagopal MR, Joranson DE, Gilson AM. Medical use, misuse,and diversion of opioids in India. Lancet. 2001;358:139–143.

[10] Devi BCR, Tang TS, Corbex M. Setting up home-based pal-liative care in countries with limited resources: a model fromSarawak, Malaysia. Ann Oncol. 2008;19:2061–2066.

[11] Krakauer EL, Wenk R, Buitrago R, Jenkins P, Scholten W. Opi-oid inaccessibility and its consequences for the field. J Pain Pal-liat Care Pharmacother. 2010;24:239–244.

[12] World Health Organization. Resolution WHA58.22 on Can-cer prevention and control. World Health Assembly, ninthplenary meeting, 25 May 2005, Committee B, third re-port. Geneva: World Health Organization; 2005. Accessed1 December 2010 from http://www.who.int/gb/ebwha/pdf files/WHA58/WHA58 22-en.pdf.

[13] United Nations. Resolution Ecosoc 2005/25 on Treatment ofpain using opioid analgesics. Thirty-sixth plenary meeting, 22July 2005. New York: United Nations; 2005.

[14] Commission on Narcotic Drugs Resolution 53/4 Promotingadequate availability of internationally controlled licit drugs formedical and scientific purposes while preventing their diversionand abuse. Report on the fifty-third session (2 December 2009and 8–12 March 2010). Economic and Social Council OfficialRecords, Supplement No. 8. New York: United Nations;2010. Accessed 1 December 2010 from http://www.unodc.org/documents/commissions/CND-Uploads/CND-53-RelatedFiles/E2010 28eV1052082.pdf.

[15] International Narcotics Control Board. Report for 2004. NewYork: United Nations; 2005. Document E/INCB/2004/1, para-graph 143.





[20] Committee on Economic, Social and Cultural Rights. Gen-eral Comment No. 14 (2000) on the right to the high-est attainable standard of health (article 12 of the In-ternational Covenant on Economic, Social and CulturalRights). New York: Economic and Social Council; 2004.E/C.12/2000/4. Accessed 1 December 2010 from http://www.unhchr.ch/tbs/doc.nsf/(symbol)/E.C.12.2000.4.En.

[21] Brennan F, Carr B, Michael C. Pain management: a fundamen-tal human right. Anesth Anal. 2007;105:205–221.

[22] Human Rights Watch. “Please, do not make us suffer any more. . .”. Access to pain treatment as a human right. Executive Sum-mary. New York; 2009.

[23] Jagwe J, Merriman A. Uganda: Delivering analgesia in ruralAfrica: opioid availability and nurse prescribing. J Pain Symp-tom Manage. 2007:33:547–551.

[24] Mosoiu D, Ryan KM, Joranson DE, Garthwaite JP. Reformof drug control policy for palliative care in Romania. Lancet.2006;367:2110–2121.

[25] Mosoiu D, Mungiu OC, Gigore B, Landon A. Romania:changing the regulatory environment. J Pain Symptom Manage.2007;33:527–532.

[26] Leon MX, DeLima L, Florez S, Torres M, Daza M, Mendoza L,et al. Improving availability of and access to opioids in Colom-bia: description and preliminary results of an action plan for thecountry. J Pain Symptom Manage. 2009;38:758–766.



[27] Senlis Council. The global unmet need for morphine.Available at: http://www.senliscouncil.net/modules/P4M/needmorphine/html met. Accessed November 2007.

[28] International Observatory on End of Life Care. Available at:www.eolc-observatory.net. Accessed November 2007.

[29] International Narcotics Control Board. Narcotic Drugs: Es-timated World Requirements for 2006; Statistics for 2004.Vienna: International Narcotics Control Board; February2006. V.05-91388. Available at: http://www.incb.org/pdf/e/tr/nar/2005/Narcotics 2005 ebook.pdf. Accessed December2007.

[30] United Nations Development Programme. Human Devel-opment Report 2004. New York: United Nations; 2006.http://hdr.undp.org/en/media/HDR06-complete.pdf. AccessedMarch 2010.

[31] Jankovic S, Vlajinacl H, Bjegovic V, Marinkovic J, Sipetic-Grujicic S, Markovic-Denic L, et al. The burden of disease andinjury in Serbia. Eur J Public Health. 2007;17:80–85.

[32] Foley KM, Wagner JL, Joranson DE, Gelband H. Paincontrol for people with cancer and AIDS. In: Disease

Control Priorities in Developing Countries. 2nd ed. 981-994. New York: Oxford University Press; Disease Con-trol Priorities in Developing Countries, KM Foley, JLWagner, DE Joranson, Eds, 2006;981–994. Available at:http://files.dcp2.org/pdf/DCP/DCP52.pdf. Accessed May 2007.

[33] Cameron A, Ewen M, Ross-Degnan D, Laing R. Medicineprices, availability, and affordability in 36 developing andmiddle-income countries: a secondary analysis. Lancet. 2008;373:240–249.

[34] De Lima L, Sweeney C, Palmer J L, Bruera E. Potent anal-gesics are more expensive for patients in developing countries:a comparative study. J Pain Palliat Care Pharmacother. 2004;18:59–70.

[35] De Lima L. Opioid availability in Latin America as a globalproblem: a new strategy with regional and national effects. J Pal-liat Med. 2004;7:97–103.

RECEIVED: 5 October 2010ACCEPTED: 27 October 2010


country need for opioids

Documents