Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/44640993

Ten-yeartrendsinhospitaladmissionsfor

adversedrugreactionsinEngland1999-2009

ArticleinJournaloftheRoyalSocietyofMedicine·June2010

DOI:10.1258/jrsm.2010.100113·Source:PubMed

CITATIONS

66

READS

65

7authors,including:

Tai-YinWu

20PUBLICATIONS199CITATIONS

SEEPROFILE

Min-HuaJen

ImperialCollegeLondon

26PUBLICATIONS447CITATIONS

SEEPROFILE

MariamMolokhia

King'sCollegeLondon

111PUBLICATIONS2,335CITATIONS

SEEPROFILE

DerekBell

ImperialCollegeLondon

167PUBLICATIONS1,952CITATIONS

SEEPROFILE

Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,

lettingyouaccessandreadthemimmediately.

Availablefrom:DerekBell

Retrievedon:18August2016

Ten-year trends in hospitaladmissions for adverse drugreactions in England 1999–2009

Tai-Yin Wu1,2,3 + Min-Hua Jen4 + Alex Bottle4 +Mariam Molokhia5 + Paul Aylin4 + Derek Bell6 + Azeem Majeed1

1 Department of Primary Care and Public Health, Imperial College London, Charing Cross Campus, London W6 8RP, UK2 Department of Family Medicine, Taipei City Hospital, Renai Branch, Taipei, Taiwan3 Institute of Preventive Medicine, National Taiwan University, Taipei, Taiwan4 Dr Foster Unit at Imperial College, Department of Primary Care and Public Health, Imperial College London, London

EC1A 9LA, UK5 Non-Communicable Disease Epidemiology Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK6 Department of Medicine, Imperial College London, Chelsea and Westminster Campus, London SW10 9NH, UK

Correspondence to: Tai-Yin Wu. E-mail: t.wu@imperial.ac.uk

Summary

Objectives Adverse drug reactions (ADR) are an important cause ofmorbidity and mortality. We analysed trends in hospital admissionsassociated with ADRs in English hospitals between 1999 and 2008.

Design Data from the Hospital Episode Statistics database were examinedfor all English hospital admissions (1999–2008) with a primary or secondarydiagnosis of an ADR recorded.

Setting All NHS (public) hospitals in England.

Main outcome measures The number of admissions and in-hospitalmortality rate with a primary (codes including ‘adverse drug reaction’, ‘drug-induced’, ‘due to drug’, ‘due to medicament’ or ‘drug allergy’) or secondarydiagnosis of ADR (ICD-10 Y40-59) were obtained and analysed. Furtheranalysis for the year 2008–2009 was performed with regard to age, gender,proportion aged >65 yrs and total bed-days.

Results Between 1999 and 2008, there were 557,978 ADR-associatedadmissions, representing 0.9% of total hospital admissions. Over this periodthe annual number of ADRs increased by 76.8% (from 42,453 to 75,076), andin-hospital mortality rate increased by 10% (from 4.3% to 4.7%). In 2008,there were 6,830,067 emergency admissions of which 75,076 (1.1%) weredrug-related. Systemic agents were most commonly implicated (19.2%),followed by analgesics (13.3%) and cardiovascular drugs (12.9%).There hasbeen a near two-fold increase in nephropathy and cardiovascularconsequences secondary to drugs and a 6.8% fall in mental and behaviouraldisorders due to drugs.

Conclusions ADRs have a major impact on public health. Our data suggestthe number of ADR admissions has increased at a greater rate than the increasein total hospital admissions; some of this may be due to improved diagnostic

DECLARATIONS

Competing interests

None declared

Funding

The Dr Foster Unit is

funded by Dr Foster

Intelligence.

Imperial College

London is grateful

for support from the

NIHR Biomedical

Research Centre

scheme and the

NIHR Collaboration

for Leadership in

Applied Health

Research & Care

(CLAHRC) scheme.

MM is funded by an

NIHR postdoctoral

award and is an

investigator for the

EU-ADR FP7 and

Serious Adverse

Events Consortium

(SAEC) projects on

adverse drug

reactions

RESEARCH

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 239

coding. However, in-hospital mortality due to ADR admissions also increasedduring the period. Our findings should prompt policymakers to implement furthermeasures to reduce ADR incidence and their associated in-hospital mortality, andmethods to improve the recording of ADRs.

Introduction

An adverse drug reaction (ADR) is an undesirableeffect of a drug beyond its anticipated therapeuticeffects occurring during clinical use,1 and is one ofthe major causes of iatrogenic disease.2 ADRscause significant morbidity and mortality and in-crease the length of hospital stays. The economicburden of ADRs on the British NHS is also high,accounting for considerable extra NHS costs.3

ADRs may be on target (where there is a clear doserelationship) and may be reversible on reducing orwithdrawing the drug,1,4 or off target (where there isuncertain dose response relationship). IdiosyncraticADRs which are largely unpredictable, can dependon the genetic susceptibility of the individual and areoften more serious (sometimes fatal). The benefit–harm relationships of ADRs are complex and shouldalways be taken into account when prescribing.

From the patients’ perspective, ADRs affectsquality of life, confidence in treatment, influenceadherence to medications, and increase costs ofcare.1 The manifestation of ADRs may also mimica disease process and, therefore, lead to furtherinvestigations and treatments.

Older people or those with multiple long-termdiseases are more susceptible to ADRs.5 As co-morbidity increases with increasing age, poly-pharmacy (use of more than four prescriptiondrugs6) is seen more frequently in older patients.7,8

Moreover, between 14 and 24% of older adultsreceive potentially inappropriate medications.9

For example, long-acting benzodiazepines have ahigh potential for severe adverse outcomes inolder patients, according to Beers criteria (a list ofinappropriate medications for the elderly).10 Poly-pharmacy increases the likelihood of drug-druginteraction and adverse drug reactions.11 ADRsthen in turn prompt the patient to seek furthermedical help and multiple primary and secondarycare visits may ensue.12

Despite the clinical impact of adverse drug re-actions, methods for monitoring them remainedlimited. Databases derived from electronic healthrecords such as the General Practice Research

Database (GPRD) or QResearch are also beingincreasingly used to monitor ADRs. Other ap-proaches include: clinical trials, prescription eventmonitoring, spontaneous reporting, and linked ad-ministrative databases. However, where an ADR israre, clinical trials lack power to detect associ-ations. Prescription event monitoring examinesrisk in newly-marketed drugs but has a relativelyshort follow-up and difficulty in selecting appro-priate controls. Spontaneous reporting is donethrough pharmacovigilance reporting such as theMHRA (Medicines and Healthcare Products Regu-latory Agency) Yellow Card Scheme in the UK,although under-reporting remains a significantissue.

Consequently, when a drug is first marketed,there is often limited information about its safetyin routine clinical use,13 although under the blacktriangle scheme new medicines are intensivelymonitored by the Commission on Human Medicines(CHM) and the MHRA. Drug safety assessmentshould be considered an integral part of everydayclinical practice,1 as well as necessity for phase IV(post-marketing surveillance) trials. Under suchpost-marketing surveillance, drugs continue to bewithdrawn from the market due to safety concerns;for example, the withdrawal of rofecoxib in 200414

due to concerns about cardiovascular safety (MI andstroke) and, more recently, efalizumab in 2009, a re-combinant humanized monoclonal antibody used totreat psoriasis, due to potential risk of progressivemultifocal leukoencephalopathy (PML).

We analysed hospital admissions associatedwith ADRs in all NHS (public) English hospitals inthe past 10 years, using the Hospital Episode Sta-tistics database. We examined the epidemiology ofhospital admissions for ADRs including temporaltrends, in-hospital mortality, the age and genderdistribution of these admissions, and differences inADR rates between individual hospitals.

Methods

Hospital Episode Statistics (HES) is an administra-tive database produced by the Department of

Ethical Approval

We have approval

from St Mary’s

Hospital Local

Research Ethics

Committee to use

routinely collected

hospital

administrative data

for research. Also,

we have been

granted permission

by the Patient

Information Advisory

Group under Section

251 of the NHS Act

2006 (originally

enacted under

Section 60 of the

Health and Social

Care Act 2001) to

use routinely

collected hospital

administrative data

to provide measures

of quality of delivery

of healthcare by

provider

Guarantors

T-YW and AM

Contributorship

T-YW extracted the

relevant ICD-10

codes, carried out

the literature review

and drafted the

manuscript. M-HJ

and AB obtained and

analysed the data

and prepared the

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113240

Health for England.15 It is available for every yearfrom 1989 onwards and comprises data gatheredlocally from each NHS hospital’s Patient Admin-istration System or Hospital Information System.4

It covers all inpatient and day-case activity deliv-ered by NHS hospitals in England, and containsdemographic, administrative and clinical infor-mation. The data-set contains over 300 fields in-cluding age, gender, admission method, primaryand secondary diagnosis fields, length of stay, andmethod of discharge (including in-hospital death).The basic unit of the data-set is the consultantepisode, covering the continuous period of timeduring which a patient was under the care of oneconsultant. Episodes of care were linked into ad-missions and those ending in transfer to anotherhospital were linked together to avoid multiplecounting. The 10th International Classification ofDiseases (ICD-10) is used for diagnosis coding.

For each financial year from 1999–2000 to 2008–2009, we obtained HES records in which there wasan ICD-10 code containing the following key-words: ‘adverse drug reaction’, ‘drug-induced’,‘due to drug’, ‘due to medicament’ or ‘drug al-lergy’. For comprehensiveness, we also searchedfor ICD-10 codes relevant to ADRs containing theword ‘immunization’, since vaccines were consid-ered as a form of medication. We then divided themain diagnostic codes into chapters based on ICD-10, as shown in Appendix 1 online (see http://jrsm.rsmjournals.com/cgi/content/full/103/6/239/DC1). We also analysed codes in the range Y40–59,which are known as ‘external cause’ codes or‘secondary diagnoses’. In this way, we excludedaccidental or intentional poisoning due to drugs.We analysed those codes that explicitly stated thatthe admission was caused by a certain drug.

HES records contain one main diagnosis codeand 13 (six up to 2000) secondary diagnosis codes.We flagged all the emergency admissions withADR-related ICD-10 codes in either the primarydiagnoses or secondary diagnoses fields. We calcu-lated the frequency of each three-digit external Ycodes and of related ADRs ICD-10 codes of eachchapter for each financial year. We excluded poss-ible duplications in coding; an admission codingboth ADR-related primary and secondary diag-noses is only counted once, as shown in Table 1.When analysed by individual ICD-10 chapters orcategories in Tables 2–5, these admissions might becounted twice or more. However, the majority of the

admissions (99.5% for year 2008–2009) only had oneADR-related ICD-10 code in their diagnosis.

For the latest year, 2008–2009, we examined thenumber of admissions by age, gender, proportionof older people (age >65 yrs), total bed-days andin-hospital mortality of each category of ADR. Weshowed the distribution of hospital-level ADRrates on a funnel plot. Statistical analysis wasperformed with SAS software (version 9.2). Forall analyses, only data for emergency (unplanned)admissions were used.

ResultsOverall hospital admissions for ADRs

Table 1 summarizes the annual number of totalhospital admissions, the number of admissions as-sociated with ADRs and in-hospital mortalityrates. In our 10-year study period, there were59,718,694 total emergency hospital admissionsand 557,978 admissions with a diagnostic code in-dicative of ADRs (0.9%). Of these, 188,024 wereprimary diagnoses and 444,780 were of ‘externalcause’. There were 26,399 in-hospital deaths inADR admissions, a case-fatality rate of 4.7%.

Ten-year trends in hospital admissions forADRs

Between 1999 and 2008, the total number of hospi-tal admissions increased by 28.3%, but the totalnumber of admissions due to ADRs increased by76.8%. The majority of the latter was due to anincrease in external cause codes. The in-hospitalmortality rate of ADR admissions increased by10%, from 4.3% to 4.7%, although this stabilizedfrom 2004 onwards (Table 1).

Tables 2 and 3 describe the burden of ADRsaccording to ICD-10 code. The number of ADRswith a primary diagnosis increased by 36.8%during this study. Nephropathy due to drugs(from 109 to 303 admissions), cardiovascular con-sequences (from 1238 to 3399 admissions) anddrug-induced lung disorders (from 64 to 157admissions) were the three fastest growing ADRsas a primary diagnosis (Table 2). Mental and be-havioural disorders due to drugs decreased 6.8%during the 10-year period, although the categorybore the highest admission burden (57,096 diag-noses in the study period). Drug-induced anaemia(−21.8%) and complications following injection,

tables and figure.

AM, PA, DB and MM

conceived the study,

mentored the other

authors and made

critical contributions

to the manuscript at

each stage

Acknowledgements

None

Reviewer

David Henry

Ten-year trends in hospital admissions for adverse drug reactions in England 1999–2009

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 241

immunization or anaesthesia (−7.0%) declinedduring the years 2004–2008.

The number of ADRs recorded in any second-ary diagnosis field increased 105.8% during thestudy period (Table 3).The three fastest growingADRs as external causes were drugs related to:gastrointestinal system (from 302 to 854 admis-sions); water-balance (from 2081 to 5789 admis-sions); and systemic agents (from 5146 to 12,667admissions). In contrast, the only reported ratewhich showed a decrease was under the categoryof bacterial vaccines (from 297 to 153 admissions).

General description for year 2008–2009

In 2008–2009, the total number of emergencyadmissions in England was 6,830,067, of which75,076 (1.1%) were drug-related (Table 1). Thetotal admission rate in England was 13,726 per100,000 person-years and the ADR admissionrate in England was 146 per 100,000 person-years(data not shown). Mental and behavioural dis-orders due to drugs (24.5%), cardiovascular con-sequences (15.1%) and complications followinginjection, immunization or anaesthesia (14.8%)were the most frequent manifestations of ADRs(Table 4). Systemic agents were the most impli-cated class (Y43, 19.2%), followed by analgesics,antipyretics and anti-inflammatory drugs (Y45,13.3%) and cardiovascular drugs (Y52, 12.9%)(Table 5). Systemic agents (Y43) were primarilyantineoplastic or immunosuppressive drugs.

The admission burden was highest for mentaland behavioural disorders due to drugs (94,579total bed-days), but in-hospital mortality rate washighest for lung disorders due to drugs (17.2%)(Table 4). The highest admission burden wasassociated with systemic agents (Y43, 112,667 totalbed-days), but the in-hospital mortality rate washighest among ADR due to agents affecting bloodconstituents (Y44, 9.4%).

Age distribution for year 2008–2009

Most hospital admissions associated with adversedrug reactions occurred in older patients. In 2008–2009, 58.5% of hospital admissions in which therewas an external ICD-10 code for ADR occurred inpeople aged 65 years and over. The median age ofADRs with an external code was 70 (Table 5). ADRsdue to bacterial vaccines (Y58) usually occurred in Ta

ble

1To

taln

umbe

rof

emer

genc

yad

mis

sion

sfo

rw

hich

ther

ew

asa

prim

ary

diag

nosi

sor

‘ext

erna

lcau

se’o

fAD

Rfr

om19

99to

2008

Year

1999

–200

020

00–2

001

2001

–200

220

02–2

003

2003

–200

420

04–2

005

2005

–200

620

06–2

007

2007

–200

820

08–2

009

Cha

nge

(%)

1999

–200

8Ch

ange

(%)

2004

–200

8

Tota

lnum

ber

ofad

mis

sion

s5,

321,

796

5,31

9,79

15,

337,

034

5,49

4,06

65,

818,

057

6,11

6,50

76,

423,

646

6,49

6,94

56,

560,

785

6,83

0,06

728

.311

.7N

umbe

rwith

drug

-indu

ced

code

s16

,434

16,2

3717

,151

16,6

4518

,307

19,6

5220

,465

20,0

0920

,685

22,4

3936

.514

.2N

umbe

rwith

exte

rnal

caus

eco

des

31,4

8632

,641

33,8

4236

,617

41,7

2244

,745

50,0

6353

,198

56,8

8363

,583

101.

942

.1To

tala

dmis

sion

sdu

eto

AD

Rs

42,4

5343

,288

45,5

1847

,455

53,2

1856

,629

61,9

3164

,536

67,8

7475

,076

76.8

32.6

In-h

ospi

talm

orta

lity

ofA

DR

ad-

mis

sion

s18

1618

7922

5924

8026

7226

7928

7830

1331

8935

3494

.631

.9

In-h

ospi

talm

orta

lity

rate

ofA

DR

adm

issi

ons

4.3

4.3

5.0

5.2

5.0

4.7

4.6

4.7

4.7

4.7

10.0

−0.5

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113242

Tabl

e2

Ann

ualn

umbe

rof

emer

genc

yad

mis

sion

sw

ith

apr

imar

ydi

agno

sis

for

anA

DR

ICD

-10

chap

ters

Des

crip

tion

1999

–20

0020

00–

2001

2001

–20

0220

02–

2003

2003

–20

0420

04–

2005

2005

–20

0620

06–

2007

2007

–20

0820

08–

2009

Cha

nge

(%)1

999–

2008

Cha

nge

(%)2

004–

2008

DD

rug-

indu

ced

anae

mia

155

138

145

162

143

316

259

178

194

247

59.4

−21.

8E

Dru

g-in

duce

dm

etab

olic

diso

rder

s85

782

482

195

410

1411

2511

8412

1813

1615

6182

.138

.8

FM

enta

land

beha

viou

ral

diso

rder

sdu

eto

drug

s59

3754

6655

1757

0359

0560

4459

8956

0353

9955

33−6

.8−8

.5

GD

rug-

indu

ced

neur

omus

cula

rdi

sord

ers

889

873

864

812

933

1042

1188

1266

1304

1578

77.5

51.4

HD

rug-

indu

ced

cata

ract

and

hear

ing

loss

2712

1010

2128

3140

2242

55.6

50.0

IC

ardi

ovas

cula

rco

nseq

uenc

esdu

eto

drug

s12

3813

8714

3814

8317

7620

7623

7424

9229

9633

9917

4.6

63.7

JD

rug-

indu

ced

lung

diso

rder

s64

4853

8880

103

101

123

110

157

145.

352

.4K

Toxi

cliv

erdi

seas

e43

844

643

848

754

949

357

853

053

361

440

.224

.5L

Der

mat

itis

due

todr

ugs

1870

1829

1866

1935

2183

2392

2719

2713

2842

3213

71.8

34.3

MD

rug-

indu

ced

imm

une

diso

rder

s19

4622

2233

0723

3822

9823

2923

6123

8024

9725

6431

.810

.1N

Nep

hrop

athy

due

todr

ugs

109

134

163

149

183

208

207

202

298

303

178.

045

.7T

Com

plic

atio

nsfo

llow

ing

inje

ctio

n,im

mun

izat

ion

oran

aest

hesi

a29

5629

3026

1426

0833

2235

9335

8933

5732

8933

4313

.1−7

.0

Tota

l16

,486

16,3

0917

,236

16,7

2918

,407

19,7

4920

,580

20,1

0220

,800

22,5

5436

.814

.2

Ten-year trends in hospital admissions for adverse drug reactions in England 1999–2009

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 243

Tabl

e3

Ann

ualn

umbe

rof

emer

genc

yad

mis

sion

sw

ith

an‘e

xter

nalc

ode’

for

anA

DR

ICD

-10

code

sD

escr

iptio

n19

99–

2000

2000

–20

0120

01–

2002

2002

–20

0320

03–

2004

2004

–20

0520

05–

2006

2006

–20

0720

07–

2008

2008

–20

09C

hang

e(%

)199

9–20

08

Cha

nge

(%)2

004–

2008

Y40

Sys

tem

ican

tibio

tics

3364

3510

3439

3719

4446

5065

5738

5976

6297

6996

108.

038

.1Y

41O

ther

syst

emic

anti-

infe

ctiv

esan

dan

tipar

asiti

cs63

571

267

675

784

396

710

6412

0512

4113

6411

4.8

41.1

Y42

Hor

mon

esan

dth

eir

synt

hetic

subs

titut

esan

dan

tago

nist

s20

8921

5420

6423

7126

3227

8530

4121

0134

9643

5210

8.3

56.3

Y43

Prim

arily

syst

emic

agen

ts51

4352

6861

6266

6368

9273

3284

4797

8911

,175

12,6

6714

6.2

72.8

Y44

Age

nts

prim

arily

affe

ctin

gbl

ood

cons

titue

nts

2939

2973

3083

3363

4003

2904

2993

2992

3055

3171

7.9

9.2

Y45

Ana

lges

ics,

antip

yret

ics

and

anti-

infla

mm

ator

ydr

ugs

4364

4385

4302

4788

5328

6078

7054

7375

7857

8747

100.

443

.9

Y46

Ant

iepi

lept

ics

and

anti-

Park

inso

nism

drug

s96

292

710

0410

2511

0612

2212

9613

1613

5414

9255

.122

.1

Y47

Sed

ativ

es,h

ypno

tics

and

anti-

anxi

ety

drug

s29

335

234

635

645

046

251

956

159

862

411

3.0

35.1

Y48

Ana

esth

etic

san

dth

erap

eutic

gase

s18

519

415

116

720

820

626

524

828

935

089

.269

.9Y

49Ps

ycho

trop

icdr

ugs,

note

lsew

here

clas

sifie

d13

7315

6314

8514

6917

9819

3519

6221

3322

5624

2276

.425

.2

Y50

Cen

tral

nerv

ous

syst

emst

imul

ants

,not

else

whe

recl

assi

fied

5264

5658

6371

111

6883

8869

.223

.9

Y51

Dru

gspr

imar

ilyaf

fect

ing

the

auto

nom

icne

rvou

ssy

stem

1698

1834

1981

2135

2559

2866

3164

3178

3331

3591

111.

525

.3

Y52

Age

nts

prim

arily

affe

ctin

gth

eca

rdio

vasc

ular

syst

em35

8740

4041

6445

2451

4759

4965

8871

0077

2385

1813

7.5

43.2

Y53

Age

nts

prim

arily

affe

ctin

gth

ega

stro

inte

stin

alsy

stem

302

338

293

321

411

457

579

584

670

854

182.

886

.9

Y54

Age

nts

prim

arily

affe

ctin

gw

ater

-bal

ance

and

min

eral

and

uric

acid

met

abol

ism

2081

2258

2613

3115

3833

4316

4946

5109

5115

5789

178.

234

.1

Y55

Age

nts

prim

arily

acts

moo

than

dsk

elet

alm

uscl

ean

dre

spir

ator

ysy

stem

292

308

270

272

329

352

383

405

416

485

66.1

37.8

Y56

Topi

cala

gent

prim

arily

affe

ctin

gsk

in/

muc

ous

mem

bran

esan

dop

th/o

t-rh

-lary

/de

ntal

drug

s

784

639

719

705

743

808

864

909

873

1020

30.1

26.2

Y57

Oth

eran

dun

spec

ified

drug

san

dm

edic

amen

ts12

9712

8013

4613

6814

4817

3821

1523

2925

4328

9212

3.0

66.4

Y58

Bac

teri

alva

ccin

es29

721

816

017

231

125

017

217

816

315

3−4

8.5

−38.

8Y

59O

ther

and

unsp

ecifi

edva

ccin

esan

dbi

olog

ical

subs

tanc

e32

926

622

221

734

035

938

033

133

641

225

.214

.8

Tota

l32

,069

33,2

8334

,536

37,5

6542

,890

46,1

2251

,681

53,8

8758

,871

65,9

8710

5.8

43.1

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113244

Tabl

e4

Num

ber

ofH

ESem

erge

ncy

adm

issi

ons

wit

ha

prim

ary

diag

nosi

sfo

ran

AD

Rin

2008

–09:

med

ian

age

ofad

mis

sion

s,ol

dag

ean

dge

nder

dist

ribu

tion

ofad

mis

sion

s,to

talb

ed-d

ays

and

mor

talit

y

ICD

-10

chap

ters

Des

crip

tion

Tota

l(%

)M

en(%

)M

edia

nag

eA

ge65

+(%

)To

talb

ed-

days

In-h

ospi

tal

mor

talit

yra

te

DD

rug-

indu

ced

anae

mia

1.1

44.1

6857

.128

987.

7E

Dru

g-in

duce

dm

etab

olic

diso

rder

s6.

944

.872

62.5

19,0

786.

2F

Men

tala

ndbe

havi

oura

ldis

orde

rsdu

eto

drug

s24

.568

.933

3.5

94,5

791.

0G

Dru

g-in

duce

dne

urom

uscu

lar

diso

rder

s7.

044

.470

60.7

32,3

896.

7H

Dru

g-in

duce

dca

tara

ctan

dhe

arin

glo

ss0.

242

.952

28.6

454

4.8

IC

ardi

ovas

cula

rco

nseq

uenc

esdu

eto

drug

s15

.146

.578

80.5

36,4

303.

8J

Dru

g-in

duce

dlu

ngdi

sord

ers

0.7

55.4

7368

.221

3017

.2K

Toxi

cliv

erdi

seas

e2.

742

.743

24.4

10,4

3011

.4L

Der

mat

itis

due

todr

ugs

14.2

41.2

6041

.942

,801

4.2

MD

rug-

indu

ced

imm

une

diso

rder

s11

.456

.469

60.9

31,5

1812

.5N

Nep

hrop

athy

due

todr

ugs

1.3

58.1

6655

.846

329.

6T

Com

plic

atio

nsfo

llow

ing

inje

ctio

n,im

mun

izat

ion

oran

aest

hesi

a14

.840

.957

30.2

26,7

172.

9

Tota

l22

,554

100.

051

.459

41.5

302,

574

4.8

Ten-year trends in hospital admissions for adverse drug reactions in England 1999–2009

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 245

Tabl

e5

Ann

ualn

umbe

rof

emer

genc

yad

mis

sion

sw

ith

an‘e

xter

nalc

ode’

for

anA

DR

in20

08–0

9:m

edia

nag

eof

adm

issi

ons,

old

age

and

gend

erdi

stri

buti

onof

adm

issi

ons,

tota

lbed

-day

san

dm

orta

lity

ICD

-10

code

sD

escr

iptio

nTo

tal(

%)

Men

(%)

Med

ian

age

Age

65+(

%)

Tota

lbed

-da

ysIn

-hos

pita

lm

orta

lity

rate

Y40

Sys

tem

ican

tibio

tics

10.6

40.5

6447

.482

,684

4.3

Y41

Oth

ersy

stem

ican

ti-in

fect

ives

and

antip

aras

itics

2.1

30.6

6449

.115

,620

2.8

Y42

Hor

mon

esan

dth

eir

synt

hetic

subs

titut

esan

dan

tago

nist

s6.

644

.567

54.6

53,0

656.

9

Y43

Prim

arily

syst

emic

agen

ts19

.242

.561

38.2

112,

667

5.4

Y44

Age

nts

prim

arily

affe

ctin

gbl

ood

cons

titue

nts

4.8

46.2

7675

.945

,865

9.4

Y45

Ana

lges

ics,

antip

yret

ics

and

anti-

infla

mm

ator

ydr

ugs

13.3

44.0

6957

.193

,656

4.4

Y46

Ant

iepi

lept

ics

and

anti-

Park

inso

nism

drug

s2.

348

.064

49.2

22,4

882.

1Y

47S

edat

ives

,hyp

notic

san

dan

ti-an

xiet

ydr

ugs

0.9

44.6

6649

.788

754.

5Y

48A

naes

thet

ics

and

ther

apeu

ticga

ses

0.5

44.9

4427

.728

612.

9Y

49Ps

ycho

trop

icdr

ugs,

note

lsew

here

clas

sifie

d3.

738

.172

61.6

36,7

673.

5Y

50C

entr

alne

rvou

ssy

stem

stim

ulan

ts,n

otel

sew

here

clas

sifie

d0.

153

.449

35.7

822

3.4

Y51

Dru

gspr

imar

ilyaf

fect

ing

the

auto

nom

icne

rvou

ssy

stem

5.4

49.6

7881

.827

,746

1.8

Y52

Age

nts

prim

arily

affe

ctin

gth

eca

rdio

vasc

ular

syst

em12

.953

.178

81.3

80,5

703.

3Y

53A

gent

spr

imar

ilyaf

fect

ing

the

gast

roin

test

inal

syst

em1.

337

.275

62.2

8790

3.9

Y54

Age

nts

prim

arily

affe

ctin

gw

ater

-bal

ance

and

min

eral

and

uric

acid

met

abol

ism

8.8

35.6

8086

.871

,974

5.4

Y55

Age

nts

prim

arily

acts

moo

than

dsk

elet

alm

uscl

ean

dre

spir

ator

ysy

stem

0.7

34.6

5130

.932

823.

3

Y56

Topi

cala

gent

prim

arily

affe

ctin

gsk

in/m

ucou

sm

embr

anes

and

opth

/ot-

rh-la

ry/d

enta

ldru

gs1.

541

.758

39.6

8061

3.8

Y57

Oth

eran

dun

spec

ified

drug

san

dm

edic

amen

ts4.

443

.369

55.6

33,8

664.

5Y

58B

acte

rial

vacc

ines

0.2

54.9

<17.

262

30.

0Y

59O

ther

and

unsp

ecifi

edva

ccin

esan

dbi

olog

ical

subs

tanc

e0.

647

.157

17.0

2070

1.5

Tota

l65

,987

100

42.9

7058

.568

0,83

34.

6

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113246

infants (median age <1), and younger patientstended to have ADRs from anaesthetics and thera-peutic gases (Y48, median age 44) or CNS stimu-lants (Y50, median age 49). Older patients (>65 yrs)tended to have ADRs due to water-balance drugs(Y54, median age 80), drugs affecting the auto-nomic nervous system (Y51, median age 78) or thecardiovascular system (Y52, median age 78).Agents primarily affecting water-balance and min-eral and uric acid metabolism (Y54) includedmineralocorticoids, diuretics, agents affecting cal-cification or uric acid metabolism, and mineralsalts.

Younger people tended to manifest ADRs asmental and behavioural disorders (median age 33)or toxic liver disease (median age 43). By contrast,those aged >65 yrs tended to have a clinical pictureof cardiovascular consequences (median age 78),lung disorders (median age 73) or metabolic dis-orders (median age 72) (Table 4).

Gender distribution for year 2008–2009

There was little gender difference in admissionswith a primary code of ADR (51.4% men) (Table 4).However, female patients predominate in admis-sions with an external cause code of ADR (42.9%men) (Table 5). Men were more likely to suffer frommental and behavioural disorders (68.9% men)and women were more likely to suffer from com-plications following injection, immunization oranaesthesia (40.9% men) (Table 4). Women appearmore likely to be affected by ADRs due to systemicanti-infectives and antiparasitics (Y41, 30.6% men),agents primarily acting on smooth and skeletalmuscle and respiratory system (Y55, 34.6% men) orwater-balance agents (Y54, 35.6% men) (Table 5).

Socioeconomic status and ADR

We analysed ADR rate per 1000 admissions bypopulation-weighted quintiles based on the Car-stairs deprivation index allocated via the postcodeto a small geographical area (‘output area’). How-ever, there was no simple socioeconomic gradientin ADR rates per 1000 admissions, which were9.9, 10.7, 10.4, 9.9 and 9.1 from quintile 1 (leastdeprived) through quintile 5 (most deprived).

Individual performance of NHS acutehospital trusts

We analysed the relationship between total emer-gency admissions and the percentage of ADR-related admissions for NHS acute hospital trusts,as shown in Figure 1. There were two trusts withunusually high rates: one with the proportion of12% (3393 admissions) and the other with 34%(1134 admissions). Both of them were not includedin the funnel plot. However, there was still widevariation in ADR-related admission percentagesbetween individual NHS trusts, with many rateslying outside the control limits despite excludingthese two trusts.

Discussion

Characteristics of this study

This study on hospital admissions for adversedrug reactions is unique in several aspects. Thestrengths of this study included its comprehen-siveness in covering all NHS hospital admissionsin England and its time trends.

Building on previous work using HES,4,16 wehave tried to be systematic in capturing all ADR-related ICD-10 codes. We had access to mortalitydata associated with hospital admissions, whichgave us a further insight into the severity of ADR-related admissions and we accessed patient-levelrather than aggregate data. A prior study con-ducted by Pirmohamed and colleagues analysedadmissions due to ADRs in two North-WestEnglish hospitals.3 However, their sample size(18,820 admissions) was substantially less thanin this nationwide survey.

Limitations of this study

The main limitation of our study is that HospitalEpisode Statistics data have the weaknesses associ-ated with all routinely collected data; namely miss-ing, incomplete or inaccurate data. Differences incoding practice between hospitals or populationscovered make comparisons at hospital level diffi-cult and will be one key explanation of the vari-ation seen in Figure 1. We were also unable to ruleout changes in coding as a result of behavioural orpolicy change as an explanation for our findings. Ifsuch variation occurs between hospitals, it is also

Ten-year trends in hospital admissions for adverse drug reactions in England 1999–2009

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 247

likely within hospitals over time, which is a poss-ible explanation for some of the trends observed.Nevertheless, HES represents one of several im-portant resources for monitoring ADRs because ofits size, national coverage and relatively standard-ized recording regulations.

We do not have data on the pre-hospital utiliza-tion of drugs (or over-the-counter drug usage), nordo we have details on the number of drugs pre-scribed or taken by patients before admission.There is no indication on the HES record whetherthe recorded ADRs were present on admission, oroccurred during a patient’s stay in hospital. Ifgiven as a primary diagnosis, in the first episode ofcare, then it is perhaps more likely that the ADRoccurred prior to admission, and if given as a sec-ondary diagnosis, then perhaps more likely tohave occurred in hospital. However, this inter-pretation is limited, as there will be some consider-able blurring between these two scenarios.

We do not have data from general practice oroutpatient visits, and we are unable to link ADRsto consultations.12 Minor ADRs which might affectthe patient’s quality of life but not to the extent ofhospital admission were not covered by our study.We were also unable to obtain data regarding riskbenefit calculations of ADRs. Hence, given thelimitations of our data, we must interpret the re-sults with caution.

Overall trends

Our study showed that the number of admissionsdue to ADRs has increased significantly over thetime period examined. The reported increase ofADR admissions by 76.8% may be due to im-proved record keeping and increased publicawareness. Population ageing also added to thistrend,17 since ADRs are most frequently encoun-tered in older people with multiple long-termconditions.

In this study, we included immunization aspart of our analysis. Complications following in-jection, immunization or anaesthesia accountedfor 31,601 admissions in the study decade, 26,717bed days in year 2008–2009, and an in-hospitalmortality rate of 2.9%. Most of the admissionsoccurred in infants.

The majority of the increase in ADR admissionswas due to increases in admissions with an ADR-related external code (105.8% change in the studyperiod). These are most likely to have developedas a result of treatments in hospital, in contrastto admissions with a primary diagnosis of ADR,which may be due to drugs prescribed in the com-munity or from a previous hospital admission.

Comparison with other studies

Pirmohamed et al. estimated that in England,ADRs were responsible for 6.5% of all emergencyhospital admissions in adults and at least 5000deaths per year.3 However, this study excludedsome categories of patients. Howard et al. sug-gested that the percentage of preventable drug-related admissions to hospitals was 3.7%.18,19 Inour study, ADR admissions accounted for only0.8–1.1% of all annual admissions for 1999–2008,and the in-hospital mortality for ADR admissionsfor 2008–2009 was 3534, which was substantiallyless than suggested in previous literature. Thereare two possible explanations for this inconsist-ency. First, this may suggest a suboptimal record-ing of ADRs in the HES database because ofunder-recognition and under-reporting of ADRs inroutine hospital activity data.16 Conversely, ourstudy is based on a comprehensive nationwidedata-set, which might also be more close to thereality than previous extrapolations from smallerscale studies.

Contradictory to some prior expectations,4 the in-troduction of the Quality and Outcome Framework

Figure 1Funnel plot of the relationship between the total emergencyadmissions and the cases of ADRs as percentage of totalemergency admissions for NHS acute hospital trusts

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113248

(QOF) on 1 April 2004 in England did not seem toaggravate the burden of ADRs. Instead, the in-hospital mortality in ADR admissions stabilized at4.7% from 2004 onwards. Drug-induced anaemiaand complications following injection, immuniz-ation or anaesthesia increased for the years 1999–2004, culminated in year 2004–2005, and thendecreased thereafter.

There is an association between increasing ageand ADR risk,20 and elderly people are more suscep-tible to ADRs. Previous studies suggested thatdiuretics and antiplatelets are often used in elderlypatients and accounted for 5.3% and 4.0%, respect-ively, of all primary care prescriptions in England.18

This study also showed similar findings, in whichagents primarily affecting water-balance (Y54) andcardiovascular system (Y52) are the most implicatedcategories in ADRs in older people.

Pirmohamed et al. reported that 59% of thepatients admitted with an ADR were female.3 Ourstudy also showed that 57% of hospital admissionswith a secondary code of ADR are female. Thepharmacokinetics and pharmacodynamics of drugmetabolism differ between the two sexes and areinfluenced by gender-specific factors such as bodybuild, hormonal transitions, or medical-seekingbehaviour.21 We cannot tell from this study if thesedifferences are due to true discrepancies in ADRoccurrence rates or are consequences of genderdifferences in prescription, consultation and usagepatterns.

As to the drug types most implicated in ADR,our study showed that systemic agents are respon-sible for one-fifth of total ADR admissions in year2008–2009, followed by analgesics and cardiovascu-lar drugs. Systemic agents are also the third fastestgrowing agent in ADR-related admissions duringthe study period. These findings are similar to pre-vious works by Waller et al.,16 but other studies alsoshowed that low-dose aspirin, diuretics, warfarin,and non-steroidal anti-inflammatory drugs werethe most commonly implicated drugs in ADR.3,8,18

Our results differed from this, and this can be ex-plained by the different classification system of thedrugs we used. We examined ICD-10 external causecodes to infer to the drug class, instead of individualgeneric name drugs. Broad groupings of drugs areused in this study, as exemplified by the systemicagent category, which includes antineoplasticdrugs, immunosuppressive agents, enzymes, anti-allergic and anti-emetic drugs.

Conclusions

The advent of modern medicine has brought hopeto many people with previously untreatable diseases,such as the use of penicillin to treat syphilis, or com-bination therapy to treat HIV infection. Nevertheless,ADRs are unwanted events following drug prescrip-tion and should be considered as part of a risk-benefitapproach when prescribing medicines, particularly invulnerable groups. Although ADRs continue to be amajor public health problem resulting in a consider-able use of NHS hospital resources and a significantnumber of deaths, their impact is often overlooked.Hospital admissions showed an apparent rise in rela-tion to ADRs; however, this may be a consequence ofimproved recording.

There are several initiatives exploring suscepti-bility to ADRs through international collabora-tions (such as FP7-funded EU-ADR, The SeriousAdverse Event Consortium, among others). Spon-taneous reporting is important in identifying pre-viously undetected ADRs, mostly type-B reactions,to allow appropriate drug regulatory action to betaken but frequently is unable to detect rare eventsuntil after years of patient use, with less than 10%reported.2 It is also a passive reporting policy andwould benefit from being linked to other data-bases. More active drug surveillance schemesmight be a potential strategy to address the highmortality associated with ADR admissions (4.7%in this study).

Other current strategies to improve drug safetyinclude stringent safety testing; nationwide imple-mentation of updated prescribing software to reduceprescribing errors;22,23 medicines reconciliation onadmission and discharge from hospital to minimizepolypharmacy;24 financial incentives for clinicians toreport ADRs to address under-recording; and fund-ing for research to decipher the complex multifacto-rial pathogenesis of idiosyncratic ADRs.25

Because of the wide range of medications avail-able, ADRs vary in manifestations and organsystem. Heightened awareness of pharmaco-vigilance by clinicians in diagnosing ADRs is crucialfor continued drug safety monitoring. Over 70%of adverse drug reactions were possibly avoidable3

and preventable through simple improvements inprescribing, e.g. increased awareness of warningprompts of possible drug interactions and high-riskpatient groups, prescribing a drug at the lowest dosenecessary to achieve the therapeutic target.

Ten-year trends in hospital admissions for adverse drug reactions in England 1999–2009

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113 249

Developed countries worldwide tend to haveprogressively older populations. In the UK, thenumber of people age 65 years and over is esti-mated to increase by 53% between 2001 and 2031.17

This increase will result in a rise in the number ofpeople at higher risk of suffering from ADRs,which are more likely to be severe in this group.26

In our study, we also noticed that most ADR in-creases occur in those over the age of 65.

In this study, systemic agents were the mostimplicated and the third fastest growing agent inADR-related admissions. This is also correlatedwith the rapid ageing of the English population,since systemic anti-cancer therapy (SACT) is tar-geted at malignancies whose risk increases withage. According to NCEPOD report 2008,27 43% ofcancer cases suffered from clinically significanttreatment-related toxicities.

Our data suggest that the number of ADR ad-missions may have increased at a greater rate thanthe increase in total hospital admissions, but it isunclear whether this could also be due to increasedrecording. Also, the percentage of ADR-related ad-missions varied considerably between individualNHS trusts. This study should prompt policy-makers toward implementing further measures toreduce ADR incidence and their associated in-hospital mortality; and to make greater use of rou-tinely collect NHS data, such as Hospital EpisodeStatistics, to monitor the epidemiology of ADRsand identify potentially avoidable aetiological fac-tors, and to improve the completeness, accuracyand standardization of coding across hospitals.

References1 Pirmohamed M, Breckenridge AM, Kitteringham NR,

Park BK. Adverse drug reactions. BMJ 1998;316:1295–82 Pirmohamed M, Park BK. Adverse drug reactions: back to

the future. Br J Clin Pharmacol 2003;55:486–923 Pirmohamed M, James S, Meakin S, et al. Adverse drug

reactions as cause of admission to hospital: prospectiveanalysis of 18,820 patients. BMJ 2004;329:15–19

4 Patel H, Bell D, Molokhia M, et al. Trends in hospitaladmissions for adverse drug reactions in England:analysis of national hospital episode statistics 1998–2005.BMC Clin Pharmacol 2007;7:9

5 Chutka DS, Takahashi PY, Hoel RW. Inappropriatemedications for elderly patients. Mayo Clin Proc2004;79:122–39

6 Viktil KK, Blix HS, Moger TA, Reikvam A. Polypharmacyas commonly defined is an indicator of limited value inthe assessment of drug-related problems. Br J ClinPharmacol 2007;63:187–95

7 Willlams CM. Using medications appropriately in olderadults. Am Fam Physician 2002;66:1917–24

8 Davies EC, Green CF, Taylor S, Williamson PR, MottramDR, Pirmohamed M. Adverse drug reactions in hospitalin-patients: a prospective analysis of 3695 patient-episodes. PLoS One 2009;4:e4439

9 Pham CB, Dickman RL. Minimizing adverse drug eventsin older patients. Am Fam Physician 2007;76:1837–44

10 Fick DM, Cooper JW, Wade WE, Waller JL, Maclean JR,Beers MH. Updating the Beers criteria for potentiallyinappropriate medication use in older adults: results of aUS consensus panel of experts. Arch Intern Med2003;163:2716–24

11 O’Mahony D, Gallagher PF. Inappropriate prescribing inthe older population: need for new criteria. Age Ageing2008;37:138–41

12 Chan DC, Hao YT, Wu SC. Characteristics of outpatientprescriptions for frail Taiwanese elders with long-termcare needs. Pharmacoepidemiol Drug Saf 2009;18:327–34

13 Murray MD, Callahan CM. Improving medication use forolder adults: an integrated research agenda. Ann InternMed 2003;139:425–9

14 Konstam MA, Weir MR, Reicin A, et al. Cardiovascularthrombotic events in controlled, clinical trials of rofecoxib.Circulation 2001;104:2280–8

15 HES Online London: Department of Health; 2009. Seehttp://www.hesonline.nhs.uk/Ease/servlet/ContentServer?siteID=1937&categoryID=658&keyword=validation

16 Waller P, Shaw M, Ho D, Shakir S, Ebrahim S. Hospitaladmissions for ‘drug-induced’ disorders in England: astudy using the Hospital Episodes Statistics (HES)database. Br J Clin Pharmacol 2005;59:213–19

17 Majeed A, Aylin P. The ageing population of the UnitedKingdom and cardiovascular disease. BMJ 2005;331:1362

18 Howard RL, Avery AJ, Slavenburg S, et al. Which drugscause preventable admissions to hospital? A systematicreview. Br J Clin Pharmacol 2007;63:136–47

19 Howard RL, Avery AJ, Howard PD, Partridge M.Investigation into the reasons for preventable drug relatedadmissions to a medical admissions unit: observationalstudy. Qual Saf Health Care 2003;12:280–5

20 Routledge PA, O’Mahony MS, Woodhouse KW. Adversedrug reactions in elderly patients. Br J Clin Pharmacol2004;57:121–6

21 Seeman MV. Gender differences in the prescribing ofantipsychotic drugs. Am J Psychiatry 2004;161:1324–33

22 Pollock M, Bazaldua OV, Dobbie AE. Appropriateprescribing of medications: an eight-step approach. AmFam Physician 2007;75:231–6

23 Kaushal R, Barker KN, Bates DW. How can informationtechnology improve patient safety and reduce medicationerrors in children’s health care? Arch Pediatr Adolesc Med2001;155:1002–7

24 NICE, NPSA. Technical patient safety solutions formedicines reconciliation on admission of adults tohospital. 2007. See http://www.nice.org.uk/nicemedia/pdf/PSG001Guidance.pdf.

25 Pirmohamed M. Genetic factors in the predisposition todrug-induced hypersensitivity reactions. AAPS J2006;8:E20–6

26 McLean AJ, Le Couteur DG. Aging biology and geriatricclinical pharmacology. Pharmacol Rev 2004;56:163–84

27 NCEPOD. For better, for worse? A review of the care ofpatients who died within 30 days of receiving systemicanti-cancer therapy. 2008. See http://www.ncepod.org.uk/2008report3/Downloads/SACT_report.pdf.

Journal of the Royal Society of Medicine

J R Soc Med 2010: 103: 239–250. DOI 10.1258/jrsm.2010.100113250