Download VOLUME 39 issue 9 September 2009

issue 9
561 Balancing patient confidentiality and the needs of
carers
563 Role of cardiac magnetic resonance imaging in
ischaemic heart disease
574 Preventing healthcare-associated infection: risks,
healthcare systems and behaviour
582 Between a rock and a hard place: exploring the
conflict between respecting the privacy of patients
and informing their carers
September 2009
617 A case of uterine tumour resembling ovarian sex
cord tumour responding to second-line, single
agent anastrazole
619 Fluoroquinolone-induced immune
thrombocytopenia: a report and review
624 Medicinal use of leeches in the texts of ancient
Greek, Roman and early Byzantine writers
628 Complications of thoracentesis
588 Understanding organ donation in the collaborative
era: a qualitative study of staff and family
experiences
629 Pulmonary toxicity associated with infliximab
therapy for ulcerative colitis
595 What we have here is a failure to communicate!
Improving communication between tertiary to
primary care for chronic heart failure patients
630 Tako-tsubo cardiomyopathy after observing
anaphylaxis
Volume 39, Issue 9, September 2009, Pages 561–632
VOLUME 39
INTERNAL
MEDICINE
JOURNAL
Volume
39
Issue
9
September
2009
ISSN 1444-0903
631 Does alcohol play a role in QT prolongation?
600 Contributors to cognitive impairment in congestive
heart failure: a pilot case–control study
606 Selection of medical patients for prophylaxis of
venous thromboembolism based on analysis of the
benefit–hazard ratio
613 Malignant fibrous histiocytoma complicating
nephrogenic systemic fibrosis post liver
transplantation
The Royal Australasian College of Physicians (RACP)
Physicians Week
Abstracts of the Australian and New Zealand Society of
Nuclear Medicine
Abstracts of the Australasian Society of Clinical
Immunology and Allergy
INTERNAL MEDICINE JOURNAL
Respecting the privacy of patients and informing their carers
Preventing iatrogenic infection: risks, healthcare systems
and behaviour
Role of cardiac magnetic resonance imaging in
ischaemic heart disease
Organ donation in the collaborative era:
staff and family experiences
Anastrazole for uterine tumour resembling
ovarian sex cord tumour
imj_v39_i9_cover_5.5mm.indd 1
9/16/2009 2:30:52 PM
WORLD MEDICINE FOR THE NEXT DECADE: 2010 to 2020
Melbourne Convention
and Exhibition Centre
Melbourne, Victoria,
Australia
www.wcim2010.com.au
REGISTRATION AND ABSTRACT SUBMISSIONS OPEN NOW
The Royal Australasian College of Physicians (RACP), the
Internal Medicine Society of Australia and New Zealand,
and the International Society of Internal Medicine invite you
to attend the World Congress of Internal Medicine 2010 in
Melbourne, Australia on 20-25 March 2010. In conjunction
with the RACP’s annual Physicians Week, both events will
ensure that this is the most significant congregation of
physicians and medical specialists to be held in the Southern
Hemisphere in 2010.
Congress Program
The Congress theme is World Medicine for the Next Decade:
2010-2020. The program will showcase excellence in
Australasian Medicine and Medical Science and will offer a
superb opportunity for physicians and medical specialists to
attend expert updates on specialty areas outside their own
specialty.
Themes running throughout the Congress will be:
t
4DJFODF3FTFBSDIBOE*OOPWBUJPO
t
1PMJDZ1FPQMFBOE1PMJUJDT
t
$MJOJDBM.FEJDJOF
t
:PVOH1FPQMFBOE)FBMUI
t
.FEJDBM&EVDBUJPO
Keynote Speakers
Professor Barry Marshall AC, together with Professor Robin
Warren, was awarded the Nobel Prize for Physiology or
Medicine in 2005. The award recognised their 1982 discovery
that a bacterium, helicobacter pylori, causes one of the most
common and important diseases of mankind, peptic ulcer
disease. In 2007 Professor Marshall was awarded the honours
PG8FTUFSO"VTUSBMJBOPGUIF:FBSBOEUIF$PNQBOJPOPGUIF
Order of Australia.
Professor Graham Brown is the Foundation Director of the
Nossal Institute for Global Health, University of Melbourne. He
has held advisory roles in health programs of organisations
such as the World Bank and the World Health Organisation.
His area of interest is Tropical Diseases, particularly malaria.
Host City – Melbourne, Australia
In the past two years Melbourne was ranked 1st & 2nd
consecutively as the World’s Most Liveable City by The
&DPOPNJTU*OUFMMJHFODF6OJU*UJTBWJCSBOUBOEDPTNPQPMJUBO
city that is host to a multitude of international events,
including the Australian Open Tennis Grand Slam and the
Australian Formula One Grand Prix. March is the best time of
year to visit Australia and New Zealand.
The new Melbourne Convention Centre is a world class,
six-star rated convention centre and the largest combined
convention centre and exhibition facility in Australia.
Visit www.wcim2010.com.au or send an email to [email protected] for further information.
ss rs
ce be
Ac em
EE m
FR ACS
R
to
Original research for physicians and paediatricians
Internal Medicine Journal
www.blackwellpublishing.com/IMJ
The Official Journal of the Adult Medicine Division of The Royal Australasian College of Physicians
(RACP)
Free access via RACP members’ website!
Manuscript Central for authors, editors and reviewers:
http://mc.manuscriptcentral.com/imj
Journal of Paediatrics and Child Health
www.blackwellpublishing.com/JPC
Official Journal of the Paediatrics and Child Health Division (RACP)
in association with the Perinatal Society of Australia and New Zealand,
the Paediatric Research Society of Australia and New Zealand and the Australasian Association of Paediatric Surgeons
Free access via RACP members’ website!
Manuscript Central for authors, editors and reviewers:
http://mc.manuscriptcentral.com/jpch
For further information and access to other Wiley-Blackwell journals online, please visit:
www.blackwellpublishing.com
html#Tax. The price includes online access from current content and all
online back files to January 1st 1997, where available. For other pricing
options, including access information and terms and conditions, please
visit www.interscience.wiley.com/journal-info.
Publisher
Internal Medicine Journal is published by Blackwell Publishing Asia
Pty Ltd
155 Cremorne Street
Richmond, Victoria, 3121
Australia
Tel: +61 3 9274 3100
Fax: +61 3 9274 3101
Email: [email protected]
Blackwell Publishing Asia Pty Ltd was acquired by John Wiley & Sons
in February 2007. Blackwell’s Programme has been merged with
Wiley’s global Scientific, Technical, and Medical business to form
Wiley-Blackwell.
Journal Customer Services
For ordering information, claims and any enquiry concerning your
journal subscription please go to interscience.wiley.com/support or
contact your nearest office.
Americas: Email: [email protected]; Tel: +1 781 388 8598 or 1 800
835 6770 (toll free in the USA & Canada).
Europe, Middle East and Africa: Email: [email protected]; Tel:
+44 (0) 1865 778315.
Asia Pacific: Email: [email protected]; Tel: +65 6511 8000.
Japan: For Japanese speaking support, email: [email protected]; Tel
(toll free): 005 316 50 480. Further Japanese customer support is also
available at www.interscience.wiley.com/support.
Production Editor
Helen Pavlatos (email: [email protected])
Information for Subscribers
Internal Medicine Journal is published in 12 issues per year. Institutional
subscriptions on prices for 2009 are: Print & Online: US$1121 (US),
US$1357 (Rest of World), @879 (Europe). Prices are exclusive of tax.
Asia-Pacific GST, Canadian GST and European VAT will be applied at the
appropriate rates. For more information on current tax rates, please go to
www3.interscience.wiley.com/aboutus/journal_ordering_and_payment.
Delivery Terms and Legal Title
Prices include delivery of print journals to the recipient’s address. Delivery
terms are Delivered Duty Unpaid (DDU); the recipient is responsible for
paying any import duty or taxes. Legal title passes to the customer on
despatch by our distributors.
Printing and Despatch
Printed in Australia by BPA Print Group.
All journals are normally despatched direct from the country in which they
are printed by surface air-lifted delivery.
Offprints
C.O.S. Printers Pte Ltd, 9 Kian Teck Crescent, Singapore 628875. Fax:
+65 6265 9074. Email: [email protected]
Back Issues
Single issues from current and recent volumes are available at the
current single issue price from [email protected]. Earlier issues
may be obtained from Periodicals Service Company, 11 Main Street,
Germantown, NY 12526, USA. Tel: +1 518 537 4700, Fax: +1 518 537 5899,
Email: [email protected]
Copyright and Photocopying
Journal compilation © 2009 Royal Australasian College of Physicians. All
rights reserved. No part of this publication may be reproduced, stored or
transmitted in any form or by any means without the prior permission in
writing from the copyright holder. Authorisation to photocopy items for
internal and personal use is granted by the copy-right holder for libraries
and other users registered with their local Reproduction Rights Organisation (RRO), e.g. Copyright Clearance Center (CCC), 222 Rosewood Drive,
Danvers, MA 01923, USA (www.copyright.com), provided the appropriate
fee is paid directly to the RRO. This consent does not extend to other kinds
of copying such as copying for general distribution, for advertising or
promotional pruposes, for creating new collective works or for resale.
Special requests should be addressed to [email protected]
IMJ.PI.SEP09
imj_v39_i9_ibc.indd 3
9/16/2009 2:29:38 PM
Honorary Advisory Board
Editor-in-Chief
Jeff Szer, Melbourne
Continuing Education/
Deputy Editor-in-Chief
Joe McCormack, Brisbane
Deputy Editor-in-Chief,
New Zealand
Zoltan Endre, Christchurch
Subspecialty Editors
Cardiology
Paul Bridgman, Christchurch
Andrew McGavigan, Adelaide
(Deputy Editor)
Clinical Pharmacology
Jenny Martin, Brisbane
Continuing Education
(Clinical Perspectives)
Christopher Pokorny, Sydney
Emergency Medicine
John Vinen, Sydney
Endocrinology
Mark McLean, Sydney
Ethics
Paul Komesaroff, Melbourne
Gastroenterology
Anne Duggan, Newcastle
Geriatric Medicine
Leon Flicker, Perth
Haematology
David Joske, Perth
Intensive Care
Michael O’Leary, Sydney
Internal Medicine
Ian Scott, Brisbane
Peter Doherty, Melbourne
Kar Neng Lai, Hong Kong
Richard Larkins, Melbourne
Greg Mundy, Nashville
Sir Gustav Nossal, Melbourne
Lawrie W. Powell, Brisbane
Nicholas Saunders, Newcastle
John Shine, Sydney
Chorh Chuan Tan, Singapore
Sir David Weatherall, Oxford
Judith Whitworth, Canberra
Manager
Virginia Savickis, Sydney
Nephrology
Zoltan Endre, Christchurch
Neurology
Peter Gates, Geelong
Editorial Office Administrator
Lorelie Willoughby, Sydney
Previous Editors-in-Chief
Nuclear Medicine
Frederick A. Khafagi, Brisbane
Occupational and
Environmental Medicine
Des Gorman, Auckland
Oncology
Damien Thomson, Brisbane
Palliative Medicine
Janet Hardy, Brisbane
Internal Medicine Journal
Edward Byrne (1999–2004)
The Australian and New Zealand
Journal of Medicine
Graham Macdonald (1989–1999)
Michael O'Rourke (1981–1989)
Akos Z. Gyory (1975–1981)
Charles Kerr (1970–1975)
The Australasian Annals of Medicine
Ronald Winton (1957–1970)
Mervyn Archdall (1952–1956)
Public Health Medicine
David Whiteman, Brisbane
Immunology and Allergy
Marianne Empson, Auckland
Respiratory Medicine
Matthew Naughton, Melbourne
Infectious Diseases
David Murdoch, Christchurch
Rheumatology
Peter Youssef, Sydney
IMJ.JEBSep09
f_i-ii_imj_v39_i9_jeb.indd i
9/16/2009 2:28:26 PM
aims and scope
The Internal Medicine Journal, formerly known as the
Australian and New Zealand Journal of Medicine, is the
official journal of the Adult Medicine Division of
The Royal Australasian College of Physicians (RACP).
Its purpose is to publish high-quality internationally
competitive peer-reviewed original medical research, both
laboratory and clinical, relating to the study and research
of human disease. Papers will be considered from all areas
of medical practice and science. The Journal also has a
major role in continuing medical education and publishes
review articles relevant to physician education. Except
where otherwise stated, articles are peer reviewed.
abstracting and indexing
This journal is indexed by Abstracts on Hygiene and
Communicable Diseases, AgBiotech News and Information,
AIDS Abstracts, Australian Medical Index, BIOBASE,
Biological Abstracts (BIOSIS), Biomedical Reference
(EBSCO), Cambridge Scientific Abstracts, Chemical
Abstracts Service, Current Contents/Clinical Medicine
(an ISI product), Derwent Biotechnology Abstracts,
EMBASE/Excerpta Medica, Environmental Sciences and
Pollution Management, Health and Safety Science Abstracts
(Online version), Helminthological Abstracts, InPharma
Weekly, International Pharmaceutical Abstracts (IPA),
Journals @ Ovid, MEDLINE, Nutrition Abstracts and
Reviews, Pharmacoeoconomics and Outcomes News,
Reactions Weekly, Science Citation Index, SCOPUS, Tropical
Diseases Bulletin,Vitis-Viticulture and Oenology Abstracts
(Online Edition),World Agricultural Economics and Rural
Sociology Abstracts, and CINAHL.
address for editorial correspondence
Editor-in-Chief, Internal Medicine Journal, The Royal
Australasian College of Physicians, 145 Macquarie Street,
Sydney, NSW 2000, Australia (tel: +61 2 9256 5431; fax:
+61 2 9256 5485) For enquiries regarding ScholarOne
Manuscripts (formerly known as ManuscriptCentral)
submissions please email [email protected]
(e.g. IMJ-0000-2009).
General enquiries should be directed to Virginia Savickis,
the Editorial Office, Internal Medicine Journal, using imj@
racp.edu.au
disclaimer
The Publisher, RACP and Editors cannot be held
responsible for errors or any consequences arising from the
use of information contained in this journal; the views and
opinions expressed do not necessarily reflect those of the
Publisher, RACP and Editors, neither does the publication
of advertisements constitute any endorsement by the
Publisher, RACP and Editors of the products advertised.
Journal compilation © 2009 Royal Australasian College
of Physicians.
For submission instructions, subscription and all other information visit www.blackwellpublishing.com/imj
This journal is available online at Wiley InterScience.Visit www.interscience.wiley.com to search the articles and register
for table of contents and email alerts.
Access to this journal is available free online within institutions in the developing world through the HINARI initiative
with the WHO. For information, visit www.healthinternetwork.org
ISSN 1444-0903 (Print)
ISSN 1445-5994 (Online)
IMJ.JEBSep09
f_i-ii_imj_v39_i9_jeb.indd ii
9/16/2009 2:28:27 PM
Share your wealth of knowledge
Submit your manuscript today
Internal Medicine Journal
The Official Journal of the Adult Medicine Division
of The Royal Australasian College of Physicians (RACP)
Internal Medicine Journal actively recruits influential
and topical material in all areas of medical practice and
science for their upcoming publications. You are invited
to submit your original medical research, whether it be
laboratory and clinical, for consideration today.
To view author guidelines visit the Journal homepage at
www.blackwellpublishing.com/imj
Edited by: Jeff Szer
Print ISSN: 1444-0903
Online ISSN: 1445-5994
Frequency: Monthly
Impact Factor (2007): 1.570
Submit your manuscript online at:
http://mc.manuscriptcentral.com/imj
®
New AFINITOR
The only therapy with proven* efficacy after
a VEGF TKI therapy in advanced RCC
• A once-daily oral mTOR inhibitor
• Powered by Phase III evidence1
• Proven to more than double progression-free
survival vs placebo after sunitinib or sorafenib
failure in advanced renal cell carcinoma (RCC)1
*In a randomised, double-blind, placebo-controlled Phase III trial.1
SEE APPROVED PRODUCT INFORMATION BEFORE PRESCRIBING. APPROVED PRODUCT INFORMATION AVAILABLE ON REQUEST. AFINITOR® (everolimus).
Indication: Treatment of patients with advanced renal cell carcinoma after failure of treatment with sorafenib or sunitinib. Dosage and administration: Adults: One 10 mg dose once daily, in either
a fasting state or after no more than a light fat-free meal. Dose adjustment may be required due to side-effects. Children: AFINITOR is not recommended for use in children or adolescents. Patients
with hepatic impairment: Dose should be reduced to 5 mg daily in patients with moderate hepatic impairment (Child-Pugh class B); not recommended in patients with severe hepatic impairment
(Child-Pugh class C). Contraindications: Hypersensitivity to everolimus, to other rapamycin derivatives or to any of the excipients. Precautions: Non-infectious pneumonitis: Cases have been
described in patients taking AFINITOR, some of these have been severe and on rare occasions, a fatal outcome was observed. In case of shortness of breath, pleural effusion, cough or dyspnoea not
due to infection or malignancy, radiologic assessment for pneumonitis is indicated. In some cases, management of pneumonitis may require dose adjustment and/or interruption, or discontinuation
of AFINITOR and/or addition of corticosteroid therapy. Infections: AFINITOR is immunosuppressive. In case of fevers and chills, signs of a potential bacterial or invasive fungal infection should be
looked for and appropriate treatment promptly instituted; treat pre-existing invasive fungal infections prior to starting treatment with AFINITOR; if a diagnosis of invasive systemic fungal infection is
made, discontinue AFINITOR and treat with appropriate antifungal therapy. Hypersensitivity reactions have been observed with everolimus and other rapamycin derivatives. Oral ulceration: Mouth
ulcers, stomatitis and oral mucositis have been seen in patients treated with AFINITOR, topical treatments are recommended, but alcohol- or peroxide-containing mouthwashes should be avoided.
Laboratory tests and monitoring: Renal function, blood glucose, and complete blood counts are recommended prior to initiation and periodically during treatment. Vaccination: Avoid use of live
vaccines. Pregnancy: Should not be given to pregnant women unless the potential benefit outweighs the potential risk to the foetus. Women of childbearing potential: Use effective contraception
methods while receiving AFINITOR, and for up to 8 weeks after ending treatment. Breast-feeding: Women taking AFINITOR should not breast feed. Fertility: Male fertility may be compromised
by treatment with AFINITOR. Avoid concurrent treatment with strong CYP3A4 inhibitors (e.g. ketoconazole, itraconazole, voriconazole, ritonavir, clarithromycin, telithromycin) and strong PgP inhibitors.
Caution with moderate inhibitors of CYP3A4 and/or PgP (e.g. erythromycin, verapamil, diltiazem, fluconazole, cyclosporin, amprenavir, fosamprenavir, aprepitant). Avoid concurrent treatment with
strong inducers of CYP3A4 or PgP (e.g. rifampicin, rifabutin). Caution with moderate inducers of CYP3A4 and PgP (e.g. St. John’s Wort (Hypericum perforatum), carbamazepine, phenobarbital,
phenytoin, efavirenz, nevirapine). Avoid grapefruit juice and grapefruit. Side effects: Very common: stomatitis, rash, fatigue, asthenia, diarrhoea, anorexia, nausea, mucosal inflammation, vomiting,
cough, infections, peripheral oedema, dry skin, epistaxis, pneumonitis, pruritus, dyspnoea, dysgeusia. Common: Headache, dry mouth, pyrexia, weight decreased, hand-foot syndrome, abdominal
pain, erythema, insomnia, dyspepsia, dysphagia, hypertension, increased daytime urination, dehydration, chest pain, haemoptysis, exacerbation of diabetes mellitus.
Uncommon: Ageusia, congestive cardiac failure, new onset diabetes mellitus. Single cases of grade 1 haemorrhages. References: 1. Motzer RJ et al, for the RECORD-1
Study Group. Lancet 2008; 372: 449–56. Novartis Pharmaceuticals Australia Pty Limited, 54 Waterloo Road, North Ryde NSW 2113. NOAF2246/IMJ. 06/09.
PBS Information: This product is not currently listed on the PBS.
®
September 2009, Volume 39, Issue 9
Editorial
561 Balancing patient confidentiality and the
needs of carers
C. White and J. Hardy
Review
563 Role of cardiac magnetic resonance imaging in
ischaemic heart disease
J. Wright and J. Bogaert
Clinical Perspectives
574 Preventing healthcare-associated infection:
risks, healthcare systems and behaviour
J. K. Ferguson
Original Articles
582 Between a rock and a hard place: exploring
the conflict between respecting the privacy of
patients and informing their carers
M. Gold, J. Philip, S. McIver and P. A. Komesaroff
588 Understanding organ donation in the
collaborative era: a qualitative study of
staff and family experiences
600 Contributors to cognitive impairment in
congestive heart failure: a pilot case–control
study
C. Beer, E. Ebenezer, S. Fenner,
N. T. Lautenschlager, L. Arnolda, L. Flicker
and O. P. Almeida
606 Selection of medical patients for prophylaxis
of venous thromboembolism based on analysis
of the benefit–hazard ratio
J. A. Millar
Brief Communications
613 Malignant fibrous histiocytoma complicating
nephrogenic systemic fibrosis post liver
transplantation
K. So, G. C. MacQuillan, L. A. Adams,
L. Delriviere, A. Mitchell, H. Moody, D. J. Wood,
R. C. Junckerstorff and G. P. Jeffrey
617 A case of uterine tumour resembling ovarian
sex cord tumour responding to second-line,
single agent anastrazole
P. Blinman and M. H. N. Tattersall
619 Fluoroquinolone-induced immune
thrombocytopenia: a report and review
C. Y. Cheah, B. De Keulenaer and M. F. Leahy
S. L. Thomas, S. Milnes and P. A. Komesaroff
History in Medicine
595 What we have here is a failure to
communicate! Improving communication
between tertiary to primary care for chronic
heart failure patients
S. Shakib, H. Philpott and R. Clark
f_v-vi_imj_v39_i9_toc.indd v
624 Medicinal use of leeches in the texts of
ancient Greek, Roman and early Byzantine
writers
N. Papavramidou and H. Christopoulou-Aletra
9/16/2009 2:28:52 PM
September 2009, Volume 39, Issue 9
Images in Medicine
628 Complications of thoracentesis
631 Does alcohol play a role in QT
prolongation?
H. J. Moriarty, T. P. Flewett and B. A. Hyslop
A. O. Soubani and M. Valdivieso
Supplements
Letters to the Editor
Clinical-scientific notes
629 Pulmonary toxicity associated with
infliximab therapy for ulcerative colitis
S. S. Heraganahally, V. Au, S. Kondru,
S. Edwards, J. J. Bowden and D. Sajkov
General correspondence
630 Tako-tsubo cardiomyopathy after observing
anaphylaxis
C. Jellis, A. Hunter, R. Whitbourn and
A. MacIsaac
f_v-vi_imj_v39_i9_toc.indd vi
The Royal Australasian College of Physicians (RACP)
Physicians Week, 17–20 May 2009, Sydney,
New South Wales, Australia
Nuclear Medicine
Abstracts of the 39th Annual Scientific Meeting of
the Australian and New Zealand Society of Nuclear
Medicine, 23–27 April 2009, Sydney Convention
Centre, Sydney, New South Wales, Australia
Clinical Immunology and Allergy
Abstracts of ASCIA 2009, 20th Annual Scientific
Meeting, Australasian Society of Clinical
Immunology and Allergy, 16–18 September 2009,
Adelaide Convention Centre, Adelaide,
South Australia, Australia
9/16/2009 2:28:52 PM
Journals & Books Online
Anatomy & Physiology • Cardiology • Dentistry • Dermatology
• Endocrine • Gastroenterology • Haematology • Immunology
• Nursing • Nutrition & Dietetics • OBGYN • Optometry &
Ophthalmology • Paediatrics • Psychiatry • Public Health
& Epidemiology • Surgery, General & Internal Medicine
www.interscience.wiley.com/journals
www.interscience.wiley.com/onlinebooks
Medicine & Healthcare
Catering for every stage of a Doctor’s career from student to practitioner and beyond…
NEW Betaferon® injection system
now available in Australia
The three major changes to the injection system are:
NEW autoinjector
• Betaject® Comfort helps patients who find manual
injections difficult
• Needle is not seen by the patient throughout the entire
injection process
• Dose delivery can be adapted for individual patient’s needs
• Needle automatically retracts once Betaferon has
been administered
• Visual and audible signals tell the patient when the injection
process is over
New BETAJECT® Comfort
NEW multipack
• 15 single-use packs containing
everything needed for one injection
• Convenient for Betaferon patients
who are travelling
A new injection system for patients with multiple sclerosis
(MS) who are using Betaferon is now available in Australia.
Betaferon was the first disease-modifying therapy introduced
for MS in Australia and has over 1,000,000 patient years
experience worldwide.
A number of changes have been made to the Betaferon
administration system, including the addition of the thinnest
needle (30G) currently used by any MS disease modifying
therapy (DMT)*.
MS rehabilitation specialist and MS Australia Medical Director
Dr Elizabeth McDonald, said Betaferon patients now have a
better way to administer their medication. “For MS patients
who have issues with injections, this system uses a much finer
needle which is not seen throughout the administration
process. MS patients tell me this delivery system is a lot easier
to administer and requires fewer steps for mixing the solution,
making necessary ongoing treatment simpler.”
NEW application system
• Thinnest needle
(30G) of all MS
disease modifying
treatments*
• New Betaferon syringe
– Larger size and
markings for increased
user-friendliness
• New three-way adapter – A new 3-way MixJect® adapter
comes with a pre-attached needle and requires fewer
steps to dissolve the Betaferon powder, shortening the
preparation process
* As of 15 July 2009
BETAFERON (interferon beta-1b) Minimum Product Information
Indications: Single clinical event suggestive of multiple sclerosis and at least two clinically silent magnetic resonance imaging (MRI) lesions characteristic of multiple
sclerosis, if alternative diagnoses have been excluded. Relapsing-remitting multiple sclerosis (greater than or equal to 2 neurological attacks/2 years with partial or
complete recovery) in ambulatory patients. Secondary progressive multiple sclerosis to reduce frequency and severity of relapses to slow disease progression. Dose:
8 x 106 IU s.c. every second day. Contraindications: Hypersensitivity to natural or recombinant interferon beta or excipients. Precautions: Depression, suicidal
ideation, seizures, myelosuppression, hypersensitivity reactions, pre-existing monoclonal gammopathy. Complete and differential white blood cell count (monitor), renal
impairment (monitor), cardiac disease, thyroid dysfunction (monitor), hepatic dysfunction, pregnancy (Category D), lactation. Adverse Effects: Injection site reactions,
flu-like symptoms, severe hypersensitivity reactions, leucopaenia, CNS disturbances incl. depression, depersonalisation and suicide ideation, elevated hepatic enzymes,
severe hepatic injury. Serum neutralising antibodies. See full PI for complete list of adverse events. Betaferon® and Betaject® are registered trademarks of the Bayer
Group, Germany. MixJect® is a registered trademark of MediMop.
Please review Product Information before prescribing. Full Product Information is
available upon request from Bayer Australia Limited, ABN 22 000 138 714, 875
Pacific Highway, Pymble, NSW 2073. AU.SM.07.2009.0064
PBS Information: Authority required: Relapsing - Remitting M.S. - INITIALLY; Ambulatory,
2 attacks in 2 years, MRI confirmation - CONTINUATION; No disability progression, compliance, tolerance.
BA396
Internal Medicine Journal 39 (2009) 561–562
E D I TO R I A L
Balancing patient confidentiality and the needs of carers
imj_2018
Decisions for healthcare professionals regarding the communication of information about patients to their carers
can be challenging, as the views of the different parties
involved may differ significantly. The balance between
respecting patient confidentiality and ‘keeping the family
happy’ can be difficult, as highlighted in the study by
Gold and colleagues in this journal.1 The study demonstrates the discrepancies between the views of patients,
carers and healthcare professionals towards the sharing
of information with carers. Many patients and their
carers stated that carers should automatically be given
information about the patient, without the need to seek
permission from the patient first. In contrast, most
healthcare professionals surveyed were of the opinion
that such information should only be shared with carers
following consent from the patient. Failure to address
and resolve these issues in individual cases may lead to
dissatisfaction and frustration on the part of patients and
their carers about the care they receive and information
they are given. Health-care professionals need therefore
to be aware of these differences.
A respect for patient autonomy and confidentiality has
been fundamental in establishing trust in the medical
profession, and underpins the modern doctor–patient
relationship. Respect for patient autonomy has been
described as an essential element of individualized,
patient-centred and ethical care. Conversely, it has been
suggested that over-emphasis may confuse and suppress
beneficial intervention.2 There are many examples in the
literature where the views of patients and carers/relatives
differ, ranging from the differential reporting of pain and
other symptoms3,4 to differences in preparedness to
consent to procedures and research.5 Furthermore, as
this study has demonstrated, while many patients were
happy for their relatives to be fully informed about their
medical care, a significant number were not. Almost a
third (30%) of patients requested that information be
given to carers only in their presence and 8% stated that
their carers should not have access to information about
them at all. In order to protect these patients, it remains
of fundamental importance that the ethical principles of
confidentiality and patient autonomy are maintained.
The involvement of carers remains of vital importance
in the management of many patients, and their need for
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
561..562
information in order to be able to provide appropriate
care cannot be ignored, especially when disease is
advanced and life-limiting. Permission from the patient
to speak to specific carers/relatives should be actively
sought by the physician early in the doctor/patient relationship, for example, as part of the initial consultation
process. Moreover, the names of those to whom information can be devolved (or not) should be documented
in the medical record, along with a discussion about the
level of information the patient is willing to share.
Patients should be reassured that they can change their
minds, withdraw or provide consent at any time for
information to be given to, or withheld from carers.
Gold and colleagues discuss how in palliative care settings the family is often automatically included in the
‘unit of care’, and information routinely given to both
patient and carer. While it is usually appropriate to share
information with carers, good medical practice dictates
that information should only be shared following consent
of the patient when the patient is mentally competent.
People with a limited prognosis still have decisionmaking capacity unless proven otherwise,6 and consent
to breach confidentiality should not be assumed. Moreover, if a patient has lost competence, healthcare decisions in many countries are the responsibility of the
nominated enduring power of attorney. The carer may
not necessarily have been appointed to this position.
Patients in any setting are potentially vulnerable,
especially when they have an advanced life-limiting
illness. They may be totally dependent on their carers/
relatives for ongoing care, in addition to their need for
ongoing support from healthcare professionals. Carers
are the ‘unsung heroes’ of today’s society. Their contribution to medical care cannot be overestimated and
their needs cannot be ignored. However, patient
autonomy and their right to confidentiality must be
paramount in order to preserve the vitally important
doctor–patient or nurse–patient relationship. Those who
want confidentiality maintained may be more reluctant
to share vital information with the professional in the
presence of a carer.
If the fundamental ethical principle of patient
autonomy is not adhered to, trust in the medical profession may diminish, to the detriment of patient care.
561
Editorial
When life expectancy is short, there is no second chance
to get it right.
2
Received 5 June 2009; accepted 16 June 2009.
doi:10.1111/j.1445-5994.2009.02018.x
3
C. White1 and J. Hardy2
1
Royal Victoria Hospital, Belfast, and Northern Ireland Hospice
Belfast, Northern Ireland
2
Mater Health Services
Brisbane, Queensland, Australia
4
5
References
1 Gold M, Philip J, McIver S, Komesaroff PA. Between a
rock and a hard place: exploring the conflict between
562
6
respecting the privacy of patients and informing their
carers. Intern Med J 2009; 39: 582–7.
Woodward VM. Caring, patient autonomy and the stigma
of paternalism. J Adv Nursing 1998; 28: 1046–52.
Higginson I, Priest P, McCarthy M. Are bereaved family
members a valid proxy for a patient’s assessment of
dying? Soc Sci Med 1994; 38: 553–7.
Clipp E, George L. Patients with cancer and their spouse
caregivers. Cancer 1992; 69: 1074–9.
White C, Hardy J, Gilshenan K, Charles M, Pinkerton R.
Randomized controlled trials of palliative care – a survey
of the views of advanced cancer patients and their
relatives. Eur J Cancer 2008; 44: 1820–8.
Addington-Hall J. Research sensitivities to palliative care
patients. Eur J Cancer Care 2002; 11: 220–4.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Internal Medicine Journal 39 (2009) 563–573
REVIEW
Role of cardiac magnetic resonance imaging in ischaemic
heart disease
imj_1884
563..573
J. Wright1,2 and J. Bogaert1
1
Department of Radiology, Gasthuisberg University Hospital, Leuven, Belgium, and 2Hearts 1st, Greenslopes Private Hospital, Brisbane, Queensland, Australia
Key words
cardiac magnetic resonance imaging,
ischaemic heart disease.
Correspondence
Jeremy Wright, Hearts 1st, Greenslopes
Private Hospital, Newdegate Street,
Greenslopes, Qld 4120, Australia.
Email: [email protected]
Abstract
Cardiac magnetic resonance imaging is a new imaging method that has much
to offer clinicians caring for patients with ischaemic heart disease. This article
describes briefly the basic principles and practical aspects of cardiac magnetic
resonance imaging, and summarizes the pathophysiology of ischaemic heart
disease. Then it discusses in detail the use of cardiac magnetic resonance
imaging for detection of coronary artery disease, and for assessment of acute
and stable coronary syndromes.
Received 1 August 2008; accepted
3 November 2008.
doi:10.1111/j.1445-5994.2008.01884.x
Introduction
Cardiovascular disease is the main cause of death in Australia and in 2002 ischaemic heart disease (IHD) caused
19.5% of all deaths. The clinical range of IHD is broad,
from being asymptomatic to stable angina, through to
sudden cardiac death, and therapeutic options vary in
expense, invasiveness and efficacy. Consequently, diagnostic imaging has become an integral part of clinical
decision-making as well as scientific endeavour. Cardiac
magnetic resonance imaging (CMR) is one of the newer
non-invasive diagnostic methods. Recent advances have
enabled CMR to come close to the goal of a complete
examination of the cardiovascular system by a single
method. It can provide relevant information on most
aspects of the heart (Table 1) – structure, global and
regional ventricular function, valve function, flow patterns, myocardial perfusion, coronary anatomy and myocardial viability, all obtained non-invasively in a single
study of 30–60 min.
This article will review the basic principles and practical
aspects of magnetic resonance imaging (MRI), then
Funding: None.
Conflict of interest: None.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
discuss in detail the use of CMR for the detection of
coronary artery disease, and for assessment of acute and
stable coronary syndromes.
Basic principles of MRI
MRI, formerly called nuclear magnetic resonance, relies
on physical properties of hydrogen nuclei (protons). These
protons, abundantly present in the human body, have an
intrinsic ‘spin’. When a patient is brought into a highstrength magnetic field, the ‘spins’ of the human body
align with the direction of the magnetic field. Application
of a radiofrequency (RF) pulse can excite the spins and
perturb their alignment. After excitation these spins gradually return to their resting state (relax), and in the process
create RF signals, which are used to create an image. The
magnitude of signal arising from the tissue is mainly
influenced by two relaxation times (T1 and T2), proton
density and movement of the protons (blood flow). The
relaxation times of the spins vary according to the local
environment, that is, the tissue that the proton is in. This
phenomenon enables the excellent soft tissue discrimination seen in MRI images. T1-weighted images exploit the
differences in T1-relaxation behaviour between tissues.
For instance, fat has a hyperintense (‘bright’) appearance,
fluid in contrast a hypointense (‘dark’) one, whereas
563
Wright & Bogaert
Table 1 Current clinical positions of the cardiac imaging methods
MRI
Myocardium
Wall thickness
Mass
Oedema
Perfusion
Ischaemia
Infarction
Stunning
Hibernation
Functional imaging
Global systolic function
Regional systolic function
Stress imaging
Diastolic function
Valvular regurgitation
Valvular stenosis
Echocardiography
MDCT
Nuclear cardiology
X-ray angiography
TTE
TEE
+++
+++
+++
++(+)
++(+)
+++
++
++(+)
++
+
+(+)
++
+
++
++
++
+
+(+)
++
+
++
++
+++
+++
+
0
+
0
0
0
++(+)
++(+)
++
++
++(+)
+
0
+
+
+
+
+++
+++
+++
++(+)
++(+)
++
++
++
+++
+++
++
+++
++
++
+++
+++
+++
++(+)
++
+(+)
+
0
0
+
++(+)
++
+++
+
+
0
++
+(+)
+(+)
+
++
++(+)
+++, excellent; ++, good; +, average; -, poor; 0, not possible. MDCT, multidetector computed tomography; MRI, magnetic resonance imaging; TEE,
transoesophageal echocardiography; TTE, transthoracic echocardiography.
myocardial tissue is isointense (‘grey’). In contrast, on
T2-weighted images fluid has a bright appearance,
whereas fat has a (less) bright appearance.
MRI contrast agents
Contrast agents are often used in assessing ischaemic
heart disease patients with CMR. The most commonly
used contrast agents contain chelates of the lanthide
metal element gadolinium with multidentate ligands
(e.g. Gd-DTPA or Gd-DOTA). These are non-specific
contrast agents that distribute throughout the extracellular space and are renally excreted in an unchanged
form. They shorten the T1 relaxation times of tissues
and therefore result in an increase in signal intensity on
T1-weighted images (lesser effect on T2). Side-effects are
very rare and these contrast agents have proved much
safer than the iodinated contrast agents used for conventional X-ray. However, there are still precautions to
be taken. Gadolinium-containing contrast agents do not
cause renal dysfunction, but should be avoided in
patients with GFR <60 mL/min per 1.73 m2 because of
the recently observed association with nephrogenic systemic fibrosis. Gadolinium should be avoided in patients
with haemolytic and sickle cell anaemia and use during
pregnancy is discouraged. Assessment of cardiac structure, global and regional myocardial function, valve
function, coronary angiography and flow quantification
can be carried out without administration of
contrast.
564
Practical aspects
A comprehensive CMR typically takes 30 min, but may
take up to 1 h when a stress study is included. The patient
must be able to fit in the magnet and lie flat and still. Most
magnets have a bore of 60 cm, but large-bore and open
magnets are available. Ideally, the patient is able to breath
hold, but real-time imaging and the use of navigator
sequences mean that excellent images can be acquired
even if the patient is unable to breath hold (patients with
New York Heart Association class IV dyspnoea can usually
be imaged provided they can lie flat). The echocardiogram
(ECG) is monitored continuously and it is safe to scan
patients within 24 h of an acute myocardial infarction
(AMI). When pharmacological stress testing is being
carried out, temporary modification of medical therapy
should be considered as for other stress testing methods
(e.g. withholding beta-blockers), but otherwise the
patient’s normal medications can continue.
Contraindications to MRI
The main contraindications to MRI relate to the presence
of metal within the patient. Non-magnetic material has a
risk of heating and electric current induction, whereas
ferromagnetic material may move in the magnetic field.
Patients with permanent pacemakers, defibrillators and
other implanted devices (neurostimulator, insulin pump,
cochlear implant etc.) should not undergo MRI. Most
prosthetic cardiac valves, coronary stents, orthopaedic
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cardiac MRI in ischaemic heart disease
implants and surgical clips are not contraindications – all
should be verified as MRI compatible before the patient
enters the scanner.
Assessment of cardiac function
One of the main strengths of CMR is the accurate assessment of ventricular volumes, mass, and global and
regional function. The accuracy, precision and reproducibility of these measurements make CMR very appealing
for primary diagnosis and follow-up studies;1 it is now
considered the reference standard.2,3
The balanced steady-state free procession (b-SSFP)
sequence is the mainstay of functional assessment, and is
not T1- or T2 weighted (signal intensity depends on the
ratio of T2/T1 in addition to flow). Thus blood, water and
fat all appear bright. Similar to all other CMR sequences,
data acquisition is synchronized to the cardiac electrical
activity using electrocardiographic triggering or gating
techniques. By acquiring data throughout the cardiac
cycle, a cine loop is reconstructed and displayed as a
movie (cine MRI). The ‘movie’ of a single slice is usually
acquired over several heartbeats; therefore, optimal
images are obtained when the rhythm is regular. If
the rhythm is irregular, non-ECG-triggered real-time
imaging is also available. It has lower spatial resolution,
but enables imaging of all patients, including those with
atrial fibrillation and very frequent ventricular ectopy.
Real-time imaging also enables assessment of patients
unable to breath hold, and those with respiratory variation in pathophysiology, for example, pericardial constriction. Using the b-SSFP sequence, the ventricles are
encompassed in a stack of 10–12 contiguous slices in
short-axis direction. The end-diastolic and end-systolic
frames are selected; then the endocardial and epicardial
contours are delineated. This allows calculation of global
functional parameters; end-diastolic volume, end-systolic
volume, stroke volume, ejection fraction, and myocardial
mass. Regional function is assessed qualitatively and
quantitatively. Qualitative evaluation of contractility is
also carried out on long-axis images and for the left
ventricle (LV) is reported using the 17-segment model
proposed by the American Society of Echocardiography.
Although echocardiography remains the reference
standard for assessment of valve leaflet morphology,
leaflets are easily seen with cine CMR techniques and
turbulent flow through stenotic or regurgitant valves is
visible. Additionally, phase-contrast velocity mapping
has proven very useful in the quantification of valvular
regurgitant fraction,4 measurement of peak velocity
through a stenotic valve, overall flow quantification
(for example Qp : Qs),5 and for assessment of diastolic
function with atrial and ventricular filling patterns.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Myocardial tagging is another CMR technique useful in
functional analysis. A grid or tag of lines on the myocardium is transiently created and these lines track the
underlying myocardial deformation. These images can be
analysed qualitatively and quantitatively, for example,
strain analysis, but the elaborate post-processing required
for quantitative analysis has largely limited it to the
research setting.
Assessment of cardiac morphology
Spin-echo sequences are useful for depiction of cardiac
morphology and produce typical ‘black-blood’ images at
a single phase of the cardiac cycle – usually mid-diastole.
These images can be T1 or T2 weighted, and can be
obtained with or without fat suppression. T1-weighted
images typically provide excellent depiction of cardiac
anatomy. In contrast, T2-weighted images provide unique
information about the free water content and are highly
useful to detect and quantify myocardial oedema, for
example, in patients with AMI.
Assessment of myocardial perfusion
‘First-pass’ imaging after i.v. injection of a small dose of
contrast is the standard CMR method of assessing myocardial perfusion (perfusion-CMR). Usually 3–5 shortaxis slices are obtained to encompass all myocardial
segments. Each slice is imaged once per heartbeat immediately after injection of gadolinium and myocardial
perfusion is qualitatively and/or semi-quantitatively
assessed.
Contrast-enhanced CMR
Contrast-enhanced CMR – also called delayed contrastenhanced, late-gadolinium enhancement, or contrastenhanced inversion recovery (CE-IR) is a key strength of
CMR and is considered the reference standard for in vivo
assessment of myocardial infarction in both the acute and
chronic phase. It has been extensively evaluated in
animal and human studies and can accurately measure
infarction to within 1 g.6–8 The technique consists of first
applying an inversion RF pre-pulse to exploit the differences in longitudinal (T1) relaxation between tissues.
Information is then acquired at a specified interval after
the pre-pulse the time interval is chosen to nullify the
signal of a specific tissue. Typically the signal of normal
myocardium is nulled (means making it dark), whereas
infarcted or scarred myocardium has a bright signal
because of the gadolinium within it. The technique is
called ‘late’ or ‘delayed’ because images are typically
obtained 10–20 min after injection of contrast. This is the
565
Wright & Bogaert
optimal time to discriminate between normal and abnormal myocardium, with the maximum difference in
gadolinium contrast concentration between normal and
abnormal tissues.
The mechanism of accumulation of contrast within
infarcted tissue is incompletely understood. Regarding
acute infarction, it is thought that myocardial cell membranes rupture, allowing gadolinium to diffuse into the
intracellular space (remembering that gadolinium is an
extracellular contrast agent). The greater distribution
area, concomitant myocardial oedema and altered contrast kinetics result in hyper-enhancement relative to the
normal myocardium. The mechanism of contrast accumulation within chronic infarcts is thought to be due to
increased interstitial space between collagen fibres, combined with slower wash-in and wash-out contrast kinetics
of infarct tissue compared with normal myocardium.
It is important to emphasize that enhancement on
contrast-enhanced CMR images is not specific for
ischaemic injury; it can also be seen in myocarditis,
infiltrative disorders (e.g. amyloid and sarcoid) and
cardiomyopathies (e.g. hypertrophic and arrhythmogenic
right-ventricular dysplasia). It is the pattern of enhancement that is used to distinguish between the different
aetiologies.
Magnet strength and CMR
Currently, a 1.5-T magnet is preferred for cardiac imaging.
Although 3-T magnets have a significant advantage
for neurological and musculoskeletal imaging, cardiac
imaging remains challenging. In particular, off-resonance
creates significant artefacts when cine imaging with
b-SSFP sequences, sometimes making the images uninterpretable. However, the increased signal-to-noise available
with 3-T magnets dramatically improves MR angiography,
delayed contrast enhancement and perfusion-CMR. Technical advances will no doubt improve the usability of 3-T
for routine cardiac applications.
Pathophysiology of ischaemic heart
disease
Broadly speaking, atherosclerotic plaque may impinge
on the coronary artery lumen slowly and progressively
(chronic stable plaque) and/or may suddenly impair
coronary flow due to a combination of plaque rupture
and thrombosis (acute coronary syndrome). Ultimately
this leads to dysfunction and/or death of myocardium.
If complete coronary artery occlusion persists for more
than 20–30 min irreversible injury occurs, starting in the
sub-endocardium and progressing towards the epicardium in the so-called wavefront of injury.9 The sub-
566
epicardial myocardial layers may survive this acute
injury, the volume of myocardium that survives depends
on many factors, including duration of ischaemia,
collateral flow, ischaemic pre-conditioning etc. Current
therapeutic strategies are focused on urgent restoration
of epicardial flow using mechanical or thrombolytic
approaches. The aim is to salvage the jeopardized but
viable myocardium in the area at risk distal to the culprit
lesion and thus to reduce adverse LV remodelling.
However, despite successful recanalization of the epicardial coronary artery with restoration of a normal flow,
not infrequently, there is severe microvascular obstruction (MVO) in the core of the infarct. In this part of the
infarct there is a complete lack of tissue perfusion due to
several pathophysiological mechanisms, including reperfusion injury and endothelial damage.10 This may occur
despite successful opening of the infarct-related epicardial coronary artery. MVO is an independent predictor of
death and adverse LV remodelling.
If occlusion is less than 15 min duration, myocyte
injury may be reversible, with prolonged but reversible
contractile dysfunction – ‘stunned myocardium’. This
may take days to weeks to normalize, although it can be
reversed with inotropic agents. When a milder degree of
ischaemia is persistent, myocytes become dysfunctional
by downregulation of energy consumption – ‘hibernating
myocardium’. If perfusion is restored before irreversible
ultrastructural changes occur, these segments may functionally return to normal.
It is important to consider the current concept of
myocardial viability – the presence of ‘life’. The current
concept contrasts with the previous definition – dysfunctional myocardium that recovers contractile function
(usually after revascularization). Whereas normal myocardial contractility implies ‘life’, the absence of contractility does not imply non-viable myocardium. There are
three subgroups of dysfunctional but viable myocardium.
Stunned or hibernating myocardial segments are
obviously viable. Additionally, a myocardial segment
with partial thickness sub-endocardial infarction and
reduced contractility is also partly viable – that is, alive.
Although this segment may not improve contractile function after successful revascularization, it still contains
viable ‘live’ myocardium.
Detection of coronary artery disease
Coronary artery imaging
MRI can be used to detect coronary artery disease morphologically or to look at the functional consequences of
coronary artery disease (ischaemia testing). Coronary
artery imaging with MRI has extensively been
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cardiac MRI in ischaemic heart disease
Figure 1 Significant stenosis in the proximal
left anterior descending coronary artery
(LAD). Coronary magnetic resonance angiography showing plaque in the proximal LAD
causing 50% stenosis (arrows a, b). The right
coronary artery (RCA) was free of disease (c). A
computed tomography coronary angiogram
confirmed calcific plaque in the LAD (d).
The corresponding short-axis slice at midventricular level showing a large perfusion
defect, seen as the dark rim in the anterior and
anteroseptal walls (e).
investigated since the early 1990s, and this application is
always regarded as one of the major strengths of CMR,
being a potential substitute for invasive coronary angiography. Despite tremendous progress in this field and the
potential to obtain high-resolution images with submillimetre spatial resolution (Fig. 1), interest has faded in
recent years. Explanations are numerous. The small size
of the coronaries (2–5 mm), their long tortuous course,
motion (respiratory, cardiac and individual artery), and
flow, necessitates long acquisition times in the order of
several minutes per coronary artery, or more than 10 min
to carry out a whole-heart MR coronary angiography
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
study. Moreover, image quality and interpretation are
often challenging in cardiac patients (compared with
healthy volunteers). Despite promising results in the
published work,11,12 coronary artery imaging is rarely
carried out in daily practice. Another equally important
explanation is the fast-growing availability of multidetector computer tomography scanners, which offer fast and
reliable imaging of the coronary arteries. The newest
scanners equipped with 320 detectors can image the
heart within a single beat, with a substantially reduced
radiation dose. This makes coronary CMR somewhat
redundant, although it remains an appealing option
567
Wright & Bogaert
when nephrotoxic contrast agents must be avoided (e.g.
severe renal dysfunction). Technical developments in the
near future may renew the interest for this exciting field
of cardiac imaging (e.g. coronary wall or plaque imaging).
Ischaemia testing
Currently two CMR techniques can be applied to functionally assess if coronary artery stenoses are flow limiting, either wall motion anlysis during pharmacologic
stress (e.g. Dobutamine), and/or first-pass perfusion
imaging (perfusion-CMR) during vasodilator stress.
Dobutamine stress MR (DSMR) uses the same principles
as Dobutamine stress echocardiography. A recent metaanalysis reported a sensitivity of 85% and specificity 86%
for the diagnosis of >50% coronary artery stenosis.13 The
main advantage DSMR has over Dobutamine stress echo
is reliable image quality; thus DSMR is preferred in
patients with suboptimal echo images. Disadvantages
include limited physical access to the patient in the MR
scanner, inability to bring a defibrillator near the magnet
and unreliable ST segment monitoring. Despite these
concerns, large studies have shown DSMR to be a safe
procedure in experienced hands.14
First-pass myocardial perfusion imaging is a robust
technique for detection of flow-limiting coronary artery
stenosis and is technically easier to carry out than DSMR.
Images are acquired at rest and during maximal pharmacological vasodilatation (adenosine or dipyridamole).
Qualitative and/or quantitative comparison of the stress
and rest images allows for accurate detection of hypoperfused myocardium (Fig. 2). A sensitivity of 91% with
specificity of 81% for the diagnosis of >50% coronary
stenosis has been reported, compared with conventional
coronary angiography.13 Specificity may improve further
if perfusion-CMR images are interpreted with delayed
contrast-enhanced images. Stress perfusion-CMR has
recently been compared with single photon emission
computed tomography (SPECT) in a large prospective
multicentre study.15 Stress perfusion-CMR was found to
be at least as good as SPECT for diagnosis of >50% coronary artery stenosis. Subgroup analysis suggested that
Figure 2 Stress perfusion-cardiac magnetic resonance imaging (CMR). A composite image showing a mid LV short-axis slice at six successive time points
during first-pass perfusion-CMR with adenosine stress. There is a significant perfusion delay in the mid inferior wall, seen as the persistent dark rim
compared with the remaining normal myocardium.
568
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cardiac MRI in ischaemic heart disease
stress perfusion-CMR was superior to non-gated SPECT,
but further research is required. In addition to diagnosis
of obstructive epicardial coronary artery disease, the high
spatial resolution of perfusion-CMR may allow detection
of sub-endocardial ischaemia – cardiac syndrome X.16 It
must be noted that these high levels of diagnostic performance are only seen in experienced centres, as stress
perfusion-CMR is prone to significant artefact.
Both DSMR and perfusion-CMR have prognostic significance in addition to diagnostic accuracy. In a large
study of patients with known or suspected coronary
artery disease, multivariate analysis found that ischaemia
detected by DSMR or stress perfusion-CMR was an independent predictor of cardiac events (hazard ratio 5.4 and
12.5, respectively).17 Patients with normal DSMR and
stress perfusion-CMR had a 3-year event-free survival of
99.2%.
Acute myocardial infarction
In general, it is safe for patients to undergo CMR within
24 h of AMI and coronary artery stents are not a contraindication. A comprehensive protocol is carried out,
with T2-weighted CMR, functional imaging, perfusionCMR, as well as contrast-enhanced CMR.
T2-weighted CMR
Myocardium with increased free water content has
increased signal intensity on T2-weighted images (i.e.
oedematous tissue appears ‘white’), allowing identification of ‘injured’ myocardium. Following infarction, myocardial oedema is usually detectable for at least 1 week
after the acute event. Increased free water content is not
confined to infarcted tissue, but is also seen in the reversibly injured or ‘jeopardized’ myocardium supplied by the
infarct-related artery. Other pathological processes such
as myocarditis, pericarditis and Tako-Tsubo cardiomyopathy can also be identified. Additionally, by comparing the
T2-weighted data with the size of myocardial infarction,
the volume of salvaged myocardium can be calculated,
which is of significant value to daily clinical practice as
well as research.18 These images can be challenging to
interpret because of image degradation (because of the
long echo times required) and a high signal ‘slow-flow’
artefact caused by stasis of blood adjacent to abnormal
myocardium.
Contrast-enhanced CMR
In the acute phase of myocardial infarction the size and
distribution of necrosis can be accurately determined. In
addition, the pattern of contrast enhancement enables
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
differentiation from non-ischaemic pathology. Several
patterns of infarction are recognized, reflecting severity
and contain prognostic information.19 Of particular
importance is the pattern of microvascular obstruction
(MVO), with transmural enhancement and a subendocardial core of hypoenhancement (Fig. 3C,D). This
core is present because of the complete lack of perfusion
at tissue level despite a patent epicardial coronary artery
and is also known as ‘no-reflow’. This phenomenon
can be identified with perfusion-CMR (Fig. 3A), but is
usually detected on ‘early’ delayed contrast-enhanced
CMR images acquired 3–5 min after contrast administration. Additionally, by comparing the contrast-enhanced
CMR images with T2 ‘oedema-weighted’ images, two
other patterns can be appreciated. Aborted myocardial
infarction shows an area of increased signal intensity on
the T2-weighted images, but no enhancement on the
contrast-enhanced CMR images. Haemorrhagic myocardial infarction is identified by a hypointense central core
on the T2-weighted images, due to blood products within
the infarction strongly decreasing the T2 relaxation time.
Complications of acute myocardial infarction
CMR offers a major advantage in the identification of
mural LV thrombus, which usually requires anticoagulation to prevent systemic embolism. Thrombi can be easily
missed with echocardiography, which has been shown to
have a sensitivity of 23% and specificity of 96%. This
compares poorly to cardiac CMR, which had sensitivity
88% and specificity 99% for detection of thrombus
(Fig. 4).20 Complications such as pericarditis, pericardial
effusion, valve dysfunction, aneurysm formation and
ventricular free wall or septal rupture can be identified in
a routine comprehensive study.
Chronic ischaemic cardiomyopathy and
myocardial viability
The presence of viable myocardium is an important factor
in the decision to refer for revascularization, particularly
when perioperative surgical risk is high. Patients with
ischaemic cardiomyopathy and angina are usually
referred for revascularization. A recent meta-analysis of
patients with ischaemic LV dysfunction reported those
with viable myocardium had a much better outcome
with successful revascularization compared with medical
therapy. In contrast, those without viability had no significant difference in outcome irrespective of treatment
strategy.21 Therefore, in the absence of angina, patients
should only be referred for surgical revascularization if
they have large areas of viable, hypokinetic and hypoperfused myocardium, or left main stenosis.
569
Wright & Bogaert
Figure 3 Cardiac magnetic resonance imaging
(CMR) in acute lateral wall myocardial infarction
with microvascular obstruction (MVO). CMR
1 week after acute lateral wall myocardial infarction (a–d). Horizontal long axis (a) and short axis
(b) contrast-enhanced CMR images showing
a large lateral myocardial infarction (white)
with a large central core of MVO (dark). The
T2-weighted image shows increased signal
intensity (white) in the infarct (C) consistent with
acute injury. The MVO is also evident on the
resting first-pass perfusion-CMR (d). Contrastenhanced CMR 4 months after the infarction
shows (e, f) resolution of the MVO and left ventricular remodelling with extensive thinning of
the lateral wall.
The preferred method of detecting viability is with
delayed contrast-enhanced CMR. As with acute infarction, ‘enhanced’ (white) myocardium is non-viable
and the remaining dark myocardium is viable. This is
the only technique able to differentiate between subendocardial and transmural infarcts and is clearly
superior to SPECT in this regard.8 The transmurality of
infarction has been shown to predict accurately functional recovery following successful revascularization in
both the acute and chronic settings. The more transmural the enhancement, the less likely is functional recovery. A study of 50 patients with chronic ischaemic LV
570
dysfunction found functional improvement in 78% of segments without hyper-enhancement and in only 1.7%
of segments with >75% transmural enhancement.22
These findings have been confirmed by other groups.
Additional accuracy for predicting functional recovery
can be obtained by combining delayed CE-IR imaging with low dose Dobutamine cine MRI, particularly
when there are segments with 25–50% transmural
enhancement.23 The third and simplest method for
assessing myocardial viability is measuring the LV wall
thickness at end diastole on the cine CMR images. A
wall thickness of <6 mm has a 92% negative predictive
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cardiac MRI in ischaemic heart disease
viability in the same study adds to the appeal of CMR.
In addition to functional recovery, revascularization of
viable myocardium may also be important for reduction
of symptoms, to prevent negative remodelling and
arrhythmia and most importantly to improve prognosis.
There are no prospective data regarding the use of
viability assessment and survival (although data are
imminent), but despite this viability assessment is part
of the current treatment guidelines.
When should I request cardiac MRI?
Figure 4 Cardiac magnetic resonance imaging (CMR) of LV apical thrombus. Contrast-enhanced CMR (a) and cine-magnetic resonance imaging
(b, end-systolic frame only) images in horizontal long axis showing a large
apical mural thrombus.
value for functional recovery after revascularization, but
wall thickness >6 mm has a poor positive predictive
value of 56%.24
It must be emphasized that to predict functional recovery one must also understand the ischaemic burden.
Indeed, it is only ischaemic hypocontractile viable myocardium that will improve contractile function after revascularization. The ability to assess perfusion, function and
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
The American College of Cardiology Foundation has
published guidelines on the appropriate use of CMR.25
Appropriate indications for CMR in patients with IHD
include instances (i) when other tests provide discordant
data or diagnostic uncertainty, (ii) when there are technically suboptimal echo images, (iii) when intracardiac
thrombus or coronary anomalies are suspected, following
myocardial infarction to determine the extent of myocardial necrosis and MVO or (iv) when coronary angiography is normal and (v) for viability assessment in patients
with LV dysfunction when revascularization is being considered. Dobutamine or vasodilator stress CMR is appropriate to assess intermediate-risk patients with chest pain
unable to exercise or with an uninterpretable ECG, and
to assess the significance of intermediate lesions detected
with coronary angiography. The European Consensus
Panel has also classified the clinical utility of CMR.26 With
respect to IHD, CMR was deemed to provide clinically
relevant information and to be clinically useful in the
following situations: assessment of global left and right
ventricular function and mass, for detection of coronary
disease with DSMR or perfusion-CMR, for angiography
of anomalous coronary arteries, for assessment of coronary artery bypass graft patency, for detection and assessment of both acute and chronic myocardial infarction,
for assessment of myocardial viability, and assessment of
ventricular thrombus.
Despite these guidelines, the role of CMR in day-today clinical practice in Australia is limited, largely
because of unavailability. Magnet access for cardiac
studies is limited in most teaching hospitals and outside
this setting Medicare item numbers are currently only
available for assessment of congenital disease and
tumour of the heart or great vessels or abnormality of
the thoracic aorta. There has also been a perception of
a lack of local expertise. There are established competency requirements for performance of CMR studies
and there is an increasing number of well-trained
(locally and overseas) practitioners in capital and other
cities around Australia.27
571
Wright & Bogaert
Conclusion
Ultimately, the role of an imaging technique is to provide
accurate information for the clinician, to minimize uncertainty in diagnostic and management decisions and optimize patient outcome. IHD has many facets and more
than one imaging technique is usually required. CMR is
increasingly available and is an attractive option because
it can accurately assess cardiac structure and function,
assess for myocardial ischaemia, quantify myocardial
infarction and its complications and assess myocardial
viability – all non-invasively and without ionizing
radiation.
10
11
12
13
References
1 Bogaert J, Bosmans H, Rademakers FE, Bellon EP,
Herregods MC, Verschakelen JA et al. Left ventricular
quantification with breath-hold MR imaging: comparison
with echocardiography. MAGMA 1995; 3: 5–12.
2 Sakuma H, Fujita N, Foo TK, Caputo GR, Nelson SJ,
Hartiala J et al. Evaluation of left ventricular volume and
mass with breath-hold cine MR imaging. Radiology 1993;
188: 377–80.
3 Grothues F, Smith GC, Moon JC, Bellenger NG,
Collins P, Klein HU et al. Comparison of interstudy
reproducibility of cardiovascular magnetic resonance
with two-dimensional echocardiography in normal
subjects and in patients with heart failure of left
ventricular hypertrophy. Am J Cardiol 2002; 90: 29–34.
4 Kozerke S, Schwitter J, Pedersen EM, Boesiger P. Aortic
and mitral regurgitation: quantification using moving
slice velocity mapping. J Magn Reson Imaging 2001; 14:
106–12.
5 Eichenberger AC, Jenni R, von Schulthess GK. Aortic
valve pressure gradients in patients with aortic valve
stenosis: quantification with velocity-encoded cine MR
imaging. Am J Roentgenol 1993; 160: 971–7.
6 Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL,
Simonetti O et al. Relationship of MRI delayed contrast
enhancement to irreversible injury, infarct age, and
contractile function. Circulation 1999; 100: 1992–2002.
7 Wu E, Judd RM, Vargas JD, Klocke FJ, Bonow RO,
Kim R. Visualization of presence, location, and
transmural extent of healed Q-wave and non-Q wave
myocardial infarction. Lancet 2001; 357: 21–8.
8 Wagner A, Mahrholdt H, Holly TA, Elliot MD,
Regenfus M, Parker M et al. Contrast-enhanced MRI and
routine single photon emission computed tomography
(SPECT) perfusion imaging for detection of
subendocardial myocardial defects: an imaging study.
Lancet 2003; 361: 374–9.
9 Reimer KA, Jennings RB. The ‘wavefront phenomenon’
of myocardial ischaemic cell death. II. Transmural
572
14
15
16
17
18
19
20
progression of necrosis within the framework of ischemic
bed size (myocardium at risk) and collateral flow. Lab
Invest 1979; 40: 633–44.
Yellon DM, Hausenloy DJ. Myocardial reperfusion injury.
N Engl J Med 2007; 357: 1121–35.
Kim YW, Danias PG, Stuber M, Flamm SD, Plein S,
Nagel E et al. Coronary magnetic resonance angiography
for the detection of coronary stenoses. N Engl J Med
2001; 345: 1863–9.
Sakuma H, Ichikawa Y, Chino S, Hirano T, Makino K,
Takeda K. Detection of coronary artery stenosis with
whole-heart coronary magnetic resonance angiography.
J Am Coll Cardiol 2006; 48: 1946–50.
Nandalur KR, Dwamena BA, Choudhri AF,
Nandalur MR, Carlos RC. Diagnostic performance of
stress cardiac magnetic resonance imaging in the
detection of coronary artery disease. J Am Coll Cardiol
2007; 50: 1343–53.
Wahl A, Paetsch I, Gollesch A, Roethemeyer S, Foell D,
Gebker R et al. Safety and feasibility of high-dose
dobutamine-atropine stress MRI for diagnosis of
myocardial ischaemia: experience in 1000 consecutive
cases. Eur Heart J 2004; 25: 1230–36.
Schwitter J, Wacker CM, van Rossum AC, Lombardi M,
Al-Saadi N, Ahlstrom H et al. MR-IMPACT: comparison of
perfusion-cardiac magnetic resonance with single photon
emission computed tomography for the detection of
coronary artery disease in a multicentre, multivendor,
randomized trial. Eur Heart J 2008; 29: 480–89.
Panting JR, Gatehouse PD, Yang GZ, Grothues F,
Firmin DN, Collins P et al. Abnormal subendocardial
perfusion in cardiac syndrome X detected by
cardiovascular magnetic resonance imaging. N Engl J Med
2002; 346: 1948–53.
Jahnke C, Nage E, Gebker R, Kokocinski T, Kelle S,
Manka R et al. Prognostic value of cardiac magnetic
resonance stress tests. Adenosine stress perfusion and
Dobutamine stress wall motion imaging. Circulation 2007;
115: 1769–76.
Friedrich MG, Abdel-Aty H, Taylor A, Schulz-Menger J,
Messroghli D, Dietz R. The salvaged area at risk in
reperfused acute myocardial infarction as visualised by
cardiovascular magnetic resonance. J Am Coll Cardiol
2008; 51: 1581–7.
De Roos A, Van Rossum AC, van der Wall E, Postema S,
Doornbos J, Matheijssen N et al. Reperfused and
nonreperfused myocardial infarction: diagnostic potential
of Gd-DTPA-enhanced MR imaging. Radiology 1989; 172:
717–20.
Srichai MB, Junor C, Rodriquez LL, Stillman AE,
Grimm RA, Lieber ML et al. Clinical, imaging, and
pathological characteristics of left ventricular thrombus:
a comparison of contrast-enhanced magnetic resonance
imaging, transthoracic-echocardiography, and
transesophageal echocardiography with surgical or
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cardiac MRI in ischaemic heart disease
21
22
23
24
25
pathological validation. Am Heart J 2006; 152:
75–84.
Allman KC, Shaw LJ, Hachamovitch R, Udelson JE.
Myocardial viability testing and impact of
revascularization on prognosis in patients with coronary
artery disease and left ventricular dysfunction: a
meta-analysis. J Am Coll Cardiol 2002; 39: 1151–8.
Kim RJ, Wu E, Rafael A, Chen EL, Parker MA,
Simonetti O et al. The use of contrast-enhanced
magnetic resonance imaging to identify reversible
myocardial dysfunction. N Engl J Med 2000; 343:
1445–53.
Wellnhofer E, Olariu A, Klein C, Grafe M, Wahl A,
Fleck E et al. Magnetic resonance low-dose dobutamine
test is superior to scar quantification for the prediction of
functional recovery. Circulation 2004; 109: 2172–4.
Schinkel AF, Bax JJ, Poldermans D, Elhendy A,
Ferrari R, Rahimtoola SH. Hibernating myocardium:
Diagnosis and Patient Outcomes. Curr Probl Cardiol 2007;
32: 375–410.
Hendel RC, Patel MR, Kramer CM, Poon M. ACCF/ACR/
SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
appropriateness criteria for cardiac computed
tomography and cardiac magnetic resonance imaging:
a report of the American College of Cardiology
Foundation/American College of Radiology, Society of
Cardiovascular Computed Tomography, Society for
Cardiovascular Magnetic Resonance, American Society of
Nuclear Cardiology, North American Society for Cardiac
Imaging, Society for Cardiovascular Angiography and
Interventions, and Society of Interventional Radiology.
J Am Coll Cardiol 2006; 48: 1475–97.
26 Pennell DJ, Sechtem UP, Higgins CB, Manning WJ,
Pohost GM, Rademakers FE et al. Clinical indications for
cardiovascular magnetic resonance (CMR): consensus
panel report. Eur Heart J 2004; 25: 1940–65.
27 Kramer CM, Budoff MJ, Fayad ZA, Ferrari VA,
Goldman C, Lesser JR et al. ACCF/AHA clinical
competence statement on vascular imaging with
computed tomography and magnetic resonance: a report
of the American College of Cardiology Foundation/
American Heart Association/American College of
Physicians Task Force on Clinical Competence and
Training. J Am Coll Cardiol 2007; 50: 1097–114.
573
Internal Medicine Journal 39 (2009) 574–581
C L I N I C A L P E R S P E CT I V E S
Preventing healthcare-associated infection: risks, healthcare
systems and behaviour
imj_2004
574..581
J. K. Ferguson1,2,3
1
Division of Medicine, John Hunter Hospital, 2Hunter Area Pathology Service, and 3University of Newcastle, Newcastle, New South Wales, Australia
Key words
healthcare-associated infection, patient safety,
quality improvement, hand hygiene, antibiotic
stewardship, risk management.
Correspondence
John Kenneth Ferguson, John Hunter Hospital,
Locked Bag 1, Newcastle Region Mail Centre,
NSW 2310, Australia. Email: john.ferguson@
hnehealth.nsw.gov.au
Received 17 November 2008; accepted 19 May
2009.
doi:10.1111/j.1445-5994.2009.02004.x
Abstract
More than 177 000 potentially preventable healthcare-associated infections
(HAIs) occur per annum in Australia with sizable attributable mortality. Organizational systems to protect against HAI in hospitals in Australia are relatively
poorly developed. Awareness and practice of infection control by medical and
other healthcare staff are often poor. These lapses in practice create significant
risk for patients and staff from HAI. Excessive patient exposure to antimicrobials is another key factor in the emergence of antibiotic-resistant bacteria and
Clostridium difficile infection. Physicians must ensure that their interactions
with patients are safe from the infection prevention standpoint. The critical
preventative practice is hand hygiene in accord with the World Health Organization 5 moments model. Improving the use of antimicrobials, asepsis and
immunization also has great importance. Hospitals should measure and feed
back HAI rates to clinical teams. Physicians as leaders, role models and educators play an important part in promoting adherence to safe practices by
other staff and students. They are also potentially effective system engineers
who can embed safer practices in all elements of patient care and promote
essential structural and organizational change. Patients and the public in
general are becoming increasingly aware of the risk of infection when entering
a hospital and expect their carers to adhere to safe practice. Poor infection
control practice will be regarded in a negative light by patients and their
families, regardless of any other manifest skills of the practitioner.
Healthcare-associated infection:
time for action
The Quality in Australian Health Care Study (QAHCS)1
estimated that 5.5% of hospital admissions were affected
by healthcare-associated infection (HAI), with an estimate at that time of 155 000 infections per annum across
Australia. These figures are consistent with other Australian estimates of between 7.7%2 and 5.7%3 and the total
estimate of 177 392 infections per annum from the Australian Commission on Safety and Quality in Healthcare
Funding: None.
Conflict of interest: None.
574
review.4 The QAHCS documented death in 4.9% of all
adverse events (including infections) and permanent disability in 13.7%. Although there has not been a systematic study of this size conducted subsequently in
Australia, there is no evidence that rates of HAI have
decreased and HAI remains a major healthcare safety
issue.1
The most common patient HAI involves the urinary
tract, respiratory tract, surgical sites, intravascular catheters and bloodstream. It is estimated that up to 70% of
HAI could be prevented if infection control procedures
were followed.5 Patients increasingly concern themselves
with risk of infection when entering a hospital and expect
their carers to adhere to safe practice. Poor infection
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Preventing healthcare-associated infection
control practice will be regarded in a negative light by
patients and their families, regardless of any other manifest skills of the practitioner.
Risks for patients
Modern healthcare generates a wide range of infection
risks for patients through practices that severely compromise host defences against infection and promote colonization by pathogenic hospital strains of bacteria. Patients
are frequently confined in crowded, difficult to clean
environments, where they may undergo invasive procedures, be fitted with prosthetic devices, and require
broad-spectrum antibiotics or immunosuppressive therapies. These conditions provide ideal opportunities for the
adaptation and spread of pathogenic microorganisms,
such as methicillin-resistant Staphylococcus aureus
(MRSA), Clostridium difficile, vancomycin-resistant
enterococci and multi-resistant Acinetobacter species.
Pathogens circulating in the general community also
enter the hospital and exploit the crowded conditions to
spread rapidly within the hospital population. Examples
include community-type MRSA, norovirus and severe
acute respiratory syndrome (SARS) virus.4 Some infections are also caused by an admitted patient’s own bacterial flora (endogenous infections), due to processes that
compromise defences against invasion and infection.
Risks to healthcare staff
Until recently many clinicians have not considered
themselves to be at risk from infection by working in
healthcare.
‘Standard Precautions’ formerly termed universal precautions (Table 1) are designed to reduce HAI risk for
both patients and staff in all healthcare settings, independent of the infectious status of a person.6 When
adhered to closely, they largely eliminate the risk of
blood-borne virus (e.g. HIV, Hepatitis B or C) transmission during healthcare without the need to document
the infection status of individual patients. There are
well-documented cases of blood-borne virus infections
in healthcare staff that have resulted from lapses in Standard Precaution practice (e.g. following avoidable needlestick injury or mucosal splash).7
The worldwide SARS epidemic provided a stark
reminder of risk from pathogenic respiratory illness. In
Canada, Hong Kong, Singapore and elsewhere, healthcare staff were at the highest risk of contracting SARS and
significant mortality resulted.8 Other respiratory illnesses
such as influenza and respiratory syncytial virus are also
frequently spread among staff by infected patients and
major morbidity may result. A recent review concludes
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
that tuberculosis remains a very important occupational
risk for healthcare staff in low and middle income countries and for staff in some high-income country facilities.
Risk appears particularly high when there is increased
exposure combined with inadequate infection control
measures.9
Organisms such as MRSA that are spread mainly by
contact (touch), colonize and infect healthcare staff with
varying ease dependent on the characteristics of the
strain. A review of 127 investigations of hospital MRSA
and the involvement of healthcare staff indicated an
average of 4.6% of healthcare staff to be carriers of MRSA
with 1 in 20 (5.1%) experiencing MRSA infection. Risk
factors for staff included chronic skin disease and poor
infection control practice. Staff were implicated as the
cause of several MRSA clusters.10 New virulent strains of
community-type MRSA have been isolated in most Australian states, causing outbreaks within families, facilities
and communities.11 MRSA infected or colonized patients
admitted to hospital increase the risk of MRSA spread
to healthcare staff and their families. A recent report
highlighted the impact of community-type MRSA on
a paediatric clinic in the United States; 16 of 45 staff
experienced skin infections and one staff member died
from MRSA infection.12
How is HAI spreads?
A conventional division specifies three modes of infectious illness dissemination – contact (direct, indirect,
common vehicle and vector), droplet and airborne
spread; however, in reality a continuum of patterns exist
for each pathogen. Table 1 provides examples of agents
predominantly spread by each mode and protective measures that are required to reduce cross-transmission.13
The hands of healthcare staff are the single most important factor in the transfer of pathogens from patient to
patient (contact spread).14 Colonized patients and staff
are the main reservoirs of hospital-adapted pathogens,
shedding these organisms into their immediate surroundings. The transfer of these pathogens (usually through
direct or indirect contact) to a normally sterile body site
or onto an invasive medical device enables an infection to
occur.
Healthcare-associated infection that result from exposure to infected aerosols (droplet or airborne transmission) are less frequent. Infections spread by the droplet
respiratory mode can also be transmitted through contact
spread as many of these infective agents remain viable on
hands and surfaces for extended periods. For instance
there is compelling evidence that respiratory syncytial
virus cross-infection in paediatrics is reduced by increasing compliance of healthcare staff with hand hygiene.15
575
Ferguson
Table 1 How are healthcare-associated infections transmitted and prevented?
Mode of transmission
Infective agents transmitted
by this mode (examples)
Protective practices
Rationale (see text as well)
Contact spread (direct/
indirect/common vehicle)
Blood-borne viruses (HIV,
Hepatitis B & C, other)
Healthcare-associated
infections, especially arising
from invasive devices or
procedures and in staff.
Standard precautions
• Aseptic technique (effective antisepsis of skin,
maximal barrier precautions during procedure,
aseptic etiquette)
• Hand washing/hand hygiene
• Use of personal protective equipment
• Safe handling/disposal of sharps/clinical waste
• Safe reprocessing of reusable equipment and
instruments
• Environmental cleaning and spills management
• Safe hospital linen and food services
Antibiotic stewardship (see text)
Immunization (see text)
Assume every individual’s blood
or body fluids are infectious
Reduce contamination of sterile
body sites during invasive
procedures
Provide additional barrier to
prevent direct exposure of staff
skin to blood/body fluids
Immunocompromised patients
are prone to certain food-borne
pathogens
Reduce antimicrobial selective
pressure
Reduce host susceptibility
Contact spread (specific
pathogens with high
epidemic potential)
Methicillin-resistant
Staphylococcus aureus, other
multi-resistant organisms
(MRO), Clostridium difficile
enteric viral infections
Transmission-based contact precautions†
• MRO screening of at-risk groups
• Isolation/cohorting of colonized/infected patients
• Impermeable gown/apron and gloves
• Enhanced cleaning and disinfection of patient
environment and equipment
Identify and contain organism
reservoir (colonized or infected
individuals)
Control of environmental
contamination
Droplet spread
Respiratory viruses, such as
influenza, Group A
streptococcus, Neisseria
meningitidis
Transmission-based droplet precautions:
• Separate unprotected contact between infected
and non-infected individuals
• Isolation or cohorting (grouping) of patients or
separation of patients
• Fluid repellent (surgical) mask
• Protective eye wear
Avoid short distance exposure to
infected respiratory droplets by
containment and distancing
Airborne spread
Pulmonary tuberculosis,
chickenpox, measles
Transmission-based airborne precautions:
• Barrier isolation in negative pressure room
• Fit-tested particulate filter (P2) mask. Staff also
fit-check the mask on each occasion a mask is
donned
• Other personal protective equipment
Healthcare staff and other
patients must be protected from
infectious fine particle (<5 mM)
aerosols that are capable of
transmitting infection at low
doses
†Active screening and isolation for methicillin-resistant Staphylococcus aureus-colonized patients/staff is not performed at some Australian sites as it is
considered to be an ineffective measure. However, all evidence-based international standards and guidelines support the practice in patients demonstrated to have moderate to high risk for carriage. Active screening for other MROs (e.g. vancomycin-resistant enterococci) is still controversial and varies
widely in practice. In large part, screening should be confined to patient populations at highest risk from morbidity (e.g. intensive care, haematology and
solid organ transplant patients).
Agents that cause respiratory infection are designated
as either spread by droplet or airborne routes (see
Table 1) largely based on epidemiological studies;
however, this division is artificial to some degree. The
process of coughing or sneezing creates droplets of
varying size that may be expelled at high velocity across
distances up to 6 m, which may facilitate distant
transfer of any respiratory infective agent. Furthermore,
in low-humidity (e.g. air conditioned) environments,
larger droplets may evaporate to form droplet nuclei
that remain airborne for extended periods. Whether
576
true airborne transmission of infection occurs depends
on such variables as the infectious dose of an agent, the
delivered dose to the recipient and what degree of
pre-existing immunity the recipient has. The microbes
that are most efficiently disseminated by the airborne
mode (e.g. tuberculosis, measles, varicella) remain
viable in droplet nuclei and have a very low infectious
dose in a susceptible individual. In low-humidity environments, influenza more usually spread by ‘droplet’
may also be spread over short distances via the airborne
route as shown from animal studies.16
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Preventing healthcare-associated infection
Prevention of HAI – approaches to
system and culture change
It is tempting to ascribe the failure to prevent HAI to
individual human factors alone – lapses and active failures by doctors and other healthcare staff; however, these
lapses occur in a healthcare environment that often fails
to facilitate safe systems of care. For instance lack of
training or credentialing in standard infection control
practices, understaffing and lack of availability of alcoholbased hand rub make adherence to safe practice less
likely. To protect patients more effectively, it is crucial
that systems of management and care in hospitals are
improved such that (i) lapses or active failures are less
likely to occur and (ii) there are safeguards to prevent
injury in the event of a lapse. A basic premise is that
humans are fallible and errors are to be expected.
Systems that provide barriers and safeguards must be
improved to reduce the capacity for human error to cause
an adverse event.17
Most existing healthcare systems still have significant
potential to create risk for patients and staff from HAI.
There are many relatively hidden and important deficiencies (latent unsafe conditions) that contribute to significant HAI risks. Table 2 catalogues Australian healthcare
systems, their status of development and examples of
latent unsafe conditions with an assessment of the HAI
risk from each of these. The risk assessment is a subjective
synthesis of the likelihood of an unsafe condition or
event coupled to the potential severity of outcome in line
with the NSW risk assessment process.18
In addition to system change, organizational culture
change, that is, the establishment of new norms of
behaviour driven by leaders in management and clinical
care who have been convinced of need for urgent change
is needed.19 Leaders must provide explicit, unequivocal
support for infection control policy and its implementation. They must ensure that all necessary enablers such as
bed-side alcohol-based hand rub and universal staff training are in place. Sufficient epidemiologists and infection
control professionals are required to effectively manage
and implement infection control surveillance, audits and
training. Once these measures are in place, there is also a
role for social marketing campaigns to healthcare staff
and patients to increase awareness of HAI and its
prevention.
Reducing HAI risk: the physician’s role
Physicians as leaders, role models, patient advocates and
educators play a crucial role in efforts to improve safety of
healthcare. A persuasive, detailed case for clinician-led
reform was made by Scott et al. recently in this journal.20
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Physicians are uniquely placed to drive clinical practice
reform that embeds specific evidence-based patient safety
practices across all relevant patient groups. As argued by
Scott et al., to achieve this, significant changes in clinical
workforce organization, teamwork, patient participation,
interventional supervision, clinical governance and
information technology for monitoring performance
are required. Systems design principles, using approaches
derived from industry, can help to make clinical care safer
and less variable and improve safety. The active involvement of clinicians in these changes is paramount.
At an individual level, physicians can make a difference
by:
• Participating in the orientation and performance management of their clinical team
• Ensuring that quality and safety issues are addressed at
clinical unit meetings
• Supporting clinical unit level programmes that facilitate quality and safety improvement and performance
audit
• Demonstrating safe practice by actions and words and
promoting safe practice among colleagues and team.
Standard and additional (transmission-based) precautions (Table 1) specify essential minimum requirements
for infection control practice.
• Increasing awareness and training about HAI
prevention at undergraduate and post-graduate levels.
For instance, requiring satisfactory compliance with hand
hygiene and other infection control measures as part of
assessment criteria for clinical vivas.
Hand hygiene
Hand hygiene is the most important element of ‘Standard
Precautions’. Microorganisms which cause infection can
be transmitted through the hands of healthcare workers
during their normal work activities. Common occurrences of this are:
• Transfer of a patient’s own microorganisms from one
body site to another
• Transfer of microorganisms from one patient to
another patient
• Transfer of microorganisms to and from the environment and healthcare equipment
• Transferring of microorganisms colonising the
healthcare worker.
Hand hygiene involves either hand washing or the use of
antiseptic alcohol-based hand rubs. Semmelweiss (1845)
famously demonstrated significant sustained reductions
in maternal postnatal mortality after enforcing hand antisepsis with a chlorinated lime solution prior to patient
care.21 Three large studies have demonstrated that multimodal programmes to achieve sustained increases in
577
Ferguson
Table 2 Healthcare systems and potential for creating risk to patients and staff from healthcare-associated infection (HAI)
System elements
Personnel management
Infection control training
Invasive procedure
credentialing
Occupational health and
safety training
Immunization
Clinical care
Standard and additional
precautions
Antibiotic stewardship
Infectious disease
management
Environmental management
Environmental cleaning and
disinfection
Built environment
(e.g. facility design)
Water
Ventilation
Waste
Food
Quality systems
Document control
Communication
HAI surveillance
Clinical pathways for common
infective syndromes
IC audit programmes
Support services
Sterilization of surgical
equipment
Sterilization and disinfection
of endoscopes
Supplier controls
Medication supply,
compounding, prescription
and administration
Existing status of
this element†
✓ Variable
✓
✓✓
✓✓
✓✓ Variable
✓
✓✓
✓ Variable
✓ Variable
✓✓✓
✓✓ Variable
✓✓✓
✓✓✓
✓ Variable
✓ Variable
✓✓
✓
✓✓
✓✓✓
✓✓
✓✓✓
✓✓
Latent unsafe conditions that
increase the risk of HAI
Risk rating‡
Staff not mandated to attend training
Staff unaware of infection control precautions
Inconsistent undergraduate training
IC requirements not integrated in to other training
Deficient asepsis during procedures and care of devices (e.g.
insertion of intravascular line)
Unsafe use/disposal of sharps
Variable reporting and management of blood-borne virus
exposures
Non-immune or staff carrying blood-borne virus allowed to
practice in situations that create patient risk (e.g. surgery)
Extreme
Variable compliance with hand hygiene and other requirements
High
Indiscriminate antibiotic exposure increases selection of
multi-resistant HAI and increases the incidence of HAI
Lack of availability or active recourse to consultation leads to
risk of death/relapse from HAI
High
Variable resources and priority given to cleaning. Variable
standards of practice. Variable training of cleaning staff.
Environmental auditing not rigorous enough. Technology;
variable adoption of more effective methods of cleaning
(e.g. new disinfection agents and modes of delivery) and audit
(e.g. use of removable surface fluorescent dye markers to
assess adequacy of cleaning)
Lack of required isolation facilities for methicillin-resistant
Staphylococcus aureus and respiratory pathogens
Poor maintenance or design elements that impede cleaning
Rare
Lack of specified respiratory isolation facilities
Rare
Adequacy of hazard analysis and critical control point plans
High
Medium
Medium
Medium
High
Medium
Low
Low
Low
Low
Informal or out-of-date guidelines remain accessible
Poorly developed communication channels among clinicians and
between management and clinicians
Increases in infection rates or outbreaks variably detected. HAI
events not validated/checked by most jurisdictions
Tolerance of variable clinical practice including delays in time to
first antibiotic dose in septic patients
Audits too infrequent, not rigorous in method; data not fed back
to clinicians
High
Medium
Rare
Low
Variable practices and training of staff
Medium
Rare
Rare
Low
Low
Medium
Medium
Medium
†
The number of ticks is a subjective assessment by the author that indicates the extent to which the system concerned has been developed and uniformly
applied across healthcare in Australia. ‡Risk stratification approach is derived from NSW Health classification (see text).18 IC, infection control.
578
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Preventing healthcare-associated infection
hand hygiene compliance have been associated with
reductions in HAI, including MRSA.22–24
A national initiative commenced in 2008 to improve
hand hygiene, focusing on education, provision of
alcohol-based hand rub at every bedside, reliable, regular
audits of compliance with effective feedback to staff and
management, and measurement of patient infection outcomes (healthcare-associated S. aureus bloodstream infections). The programme is modelled on the five moments
for hand hygiene programme developed by the World
Health Organization. A central message is to clean your
hands before and after every patient contact.
In observational audits, medical staff are often demonstrated to have the poorest hand hygiene adherence and
thereby expose their patients to significant risk. As
leaders, mentors, educators and patient advocates,
physicians must urgently adopt a best practice of hand
hygiene throughout their practice settings. System
aspects include making sure that hand alcohol-based
hand rub is available for use at each bed-side and practice
setting. Just as the culture change that normalized the
use of seat belts was essential to improvements in road
safety, sustained changes in compliance with hand
hygiene will only arise out medical support for a pervasive organizational (and perhaps regulatory) approach to
culture change.
Other elements of the hospital ward round may
create significant risk to patients and staff due to lack of
compliance with Standard Precautions. In particular,
portable equipment (e.g. stethoscope) taken to the
bedside must be cleaned or disinfected prior to contact
with a patient or their environs. A good system of care
on ward rounds is to provide a separate ward round
trolley equipped with alcohol-based hand rub and large
alcohol-impregnated wipes for disinfecting stethoscopes
and other equipment. Such a trolley can also hold
the patient clinical files and provide a surface for
writing, avoiding cross-contamination with the patient
environment.
Asepsis during invasive procedures
Asepsis encompasses techniques, including disinfection,
that reduce the potential for microorganisms to contaminate sterile body sites during invasive procedures. As an
example, studies of central intravascular lines document
significant reduction in risk from infection when optimal
aseptic practices are systematically adopted. A care
‘bundle’ that includes performance of hand hygiene by
the operator prior to insertion, application of effective
skin antiseptic, allowing it sufficient time to work,
wearing sterile protective apparel, and using a large
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
sterile drape for the insertion site virtually eliminates
intensive care central line-associated bloodstream
infection.25
Antimicrobial stewardship
Prior patient exposure to antimicrobials is a key risk
factor for colonization and infection due to antibioticresistant bacteria and C. difficile infection. These infections
usually add to the infectious burden rather than just
replacing existing cases of infection caused by less resistant pathogens. Evidence from community and hospital
practice shows that use of systemic antimicrobials is often
indiscriminate or ineffectively targeted against the likely
or proven pathogen.
Antimicrobial stewardship programmes attempt to
improve prescribing to reduce unnecessary use and
promote effective directed antibiotic treatment in line
with guidelines and demonstrated incidence of antibiotic resistance. Successful stewardship programmes
have been shown to reduce not only resistance rates in
hospitalized patients but also morbidity, mortality and
cost.26 The Australian Commission on Safety and
Quality in Healthcare has recently established a project
to improve antibiotic stewardship in Australian
hospitals.27
Key practice points include27,28
• Potentially septic patients need appropriate investigation prior to antibiotic treatment – at least two blood
culture sets from different peripheral sites, other microbiology as indicated
• Empiric antimicrobial choice and dose for septic
patients should be based on recommendations from
Therapeutic Guidelines: Antibiotic or local Infectious
Diseases/Microbiology expert advice – inadequate initial
therapy is a demonstrated risk factor for adverse
outcomes
• Indications for antimicrobial treatment and duration
should always be documented
• Patients who are on antimicrobials need regular evaluation to determine: need for ongoing treatment and/or
need to target (or direct) treatment against a demonstrated pathogen (select alternative agent, consider
correct dose, switch to oral, modify treatment plan
including duration of treatment)
• Follow recommended surgical antibiotic prophylaxis
(right drug, right dose, right timing – administer within
60 min of procedure commencement, no post-operative
doses)
• Evidence-based computerized decision support
systems facilitate better prescribing and lower bacterial
resistance.
579
Ferguson
Immunization
Immunization of healthcare staff helps protect the individual and also reduces risk from vaccine preventable
disease in patients. For instance high uptake rates of
annual influenza immunization by healthcare staff in
aged care facilities has been shown to reduce mortality in
their elderly patients.29
All healthcare staff with direct patient contact need to
ensure that their immune status is optimized for hepatitis
B, tuberculosis, measles, chickenpox, influenza and pertussis. Medical staff should visit their staff health service
annually to update their immunization and have their
immune status checked as required.
The Australian Immunisation handbook defines many
situations in which patients who are over 65 years and
younger patients with various chronic medical conditions
should receive additional regular immunizations. These
include patients with splenectomy or hyposplenism.
Various surveys of these patients indicate that compliance
with immunization guidelines is poor.30 All physicians
should implement systems of care to identify their at-risk
patients to enable opportunistic immunization as
recommended.
Surveillance
Measurement of the incidence of major types of HAI is an
essential component of control programmes.1 Bloodstream infections, surgical site infections, intensive care
infections, infections and colonizations due to multiresistant organisms are usually documented by routine
surveillance systems and reported to State and National
bodies.
Clinical teams should receive regular feedback about
HAI, antibiotic resistance and usage in their patients.
Conclusions
Current levels of HAI in Australian hospitals are unacceptable and lead to preventable morbidity. Physicians
can drive widespread system and practice change towards
safer care using existing knowledge about quality
improvement. The impact of such changes will be evident
from HAI surveillance data and will serve to increase the
community’s trust in the healthcare system.
Acknowledgements
Thank you to Dr Rod Givney (Hunter Area Pathology
Service, Newcastle), Dr Kim Hill (Hunter New England
Health, Newcastle), Ms Sandy Berenger (Infection Prevention and Control, Hunter New England Health, New-
580
castle), Dr Michael Boyle (John Hunter Hospital, Hunter
New England Health, Newcastle), Dr Craig Boutlis
(Wollongong Hospital) and Dr Celia Cooper (Women’s
and Children’s Hospital, Adelaide) who provided helpful
suggestions.
References
1 Wilson RM, Runciman WB, Gibberd RW, Harrison BT,
Newby L, Hamilton JD. The Quality in Australian Health
Care Study. Med J Aust 1995; 163: 458–71.
2 McLaws ML, Gold J, King K, Irwig LM, Berry G. The
prevalence of nosocomial and community-acquired
infections in Australian hospitals. Med J Aust 1988; 149:
582–90.
3 Graves N, Weinhold D, Tong E, Birrell F, Doidge S,
Ramritu P et al. Effect of healthcare-acquired infection on
length of hospital stay and cost. Infect Control Hosp
Epidemiol 2007; 28: 280–92.
4 Cruickshank M, Ferguson JK. Reducing Harm to Patients
from Healthcare Associated Infection: The Role of Surveillance.
Sydney, NSW: Australian Commission on Safety and
Quality in Healthcare; 2008.
5 Harbarth S, Sax H, Gastmeier P. The preventable
proportion of nosocomial infections: an overview of
published reports. J Hosp Infect 2003; 54: 258–66. quiz
321.
6 Australian Government DoHaA. Infection Control
Guidelines. Table 2.1 Standard precautions for infection
control in health care settings. Canberra, ACT:
Department of Health and Ageing; 2004.
7 Do AN, Ciesielski CA, Metler RP, Hammett TA, Li J,
Fleming PL. Occupationally acquired human
immunodeficiency virus (HIV) infection: national case
surveillance data during 20 years of the HIV epidemic in
the United States. Infect Control Hosp Epidemiol 2003; 24:
86–96.
8 World Health Organization. Consensus Document on the
Epidemiology of Severe Acute Respiratory Syndrome
(SARS). 2003; 14 [cited 2008 Nov 1]. Available from:
URL: http://www.who.int/entity/csr/sars/WHOconsensus.
pdf
9 Menzies D, Joshi R, Pai M. Risk of tuberculosis infection
and disease associated with work in health care settings.
Int J Tuberc Lung Dis 2007; 11: 593–605.
10 Albrich WC, Harbarth S. Health-care workers: source,
vector, or victim of MRSA? Lancet Infect Dis 2008; 8:
289–301.
11 Nimmo GR, Coombs GW. Community-associated
methicillin-resistant Staphylococcus aureus (MRSA)
in Australia. Int J Antimicrob Agents 2008; 31:
401–10.
12 Carpenter LR, Kainer M, Woron A, Schaffner W,
Jones TF. Methicillin-resistant Staphylococcus aureus and
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Preventing healthcare-associated infection
13
14
15
16
17
18
19
20
21
22
23
24
25
26
skin infections among personnel at a pediatric clinic. Am
J Infect Control 2008; 36: 665–7.
Siegal HD, Rhinehart RN, Jackson M, Chiarello L,
Healthcare Infection Control Practices Advisory
Committee. 2007 Guidelines for Isolation Precautions:
Preventing Transmission of Infectious Agents in
Healthcare Settings. 2007 [cited 2009 Jun 1]. Available
from: URL: http://www.cdc.gov/neidod/dhqp/pdf/
guidelines/Isolation2007.pdf
Trampuz A, Widmer AF. Hand hygiene: a frequently
missed lifesaving opportunity during patient care. Mayo
Clin Proc 2004; 79: 109–16.
Hall CB. Nosocomial respiratory syncytial virus
infections: the ‘Cold War’ has not ended. Clin Infect Dis
2000; 31: 590–6.
Brankston G, Gitterman L, Hirji Z, Lemieux C,
Gardam M. Transmission of influenza A in human
beings. Lancet Infect Dis 2007; 7: 257–65.
Reason J. Human error: models and management. BMJ
2000; 320: 768–70.
New South Wales Health. Severity assessment code
matrix. 2005 [cited 2008 Nov 1]. Available from: URL:
http://www.health.nsw.gov.au/pubs/2005/
sac_matrix.html
Kotter JP. Leading Change. Boston (MA): Harvard
Business School Press; 1996.
Scott IA, Poole PJ, Jayathissa S. Improving quality and
safety of hospital care: a reappraisal and an agenda for
clinically relevant reform. Intern Med J 2008; 38: 44–55.
Pittet D, Boyce JM. Hand hygiene and patient care:
pursuing the Semmelweis legacy. Lancet Infect Dis 2001;
9–20.
Grayson ML, Jarvie LJ, Martin R, Johnson PD,
Jodoin ME, McMullan C et al. Significant reductions in
methicillin-resistant Staphylococcus aureus bacteraemia and
clinical isolates associated with a multisite, hand hygiene
culture-change program and subsequent successful
statewide roll-out. Med J Aust 2008; 188: 633–40.
Pittet D, Hugonnet S, Harbarth S, Mourouga P,
Sauvan V, Touveneau S et al. Effectiveness of a
hospital-wide programme to improve compliance with
hand hygiene. Infection Control Programme. Lancet.
2000; 356: 1307–12.
Johnson PD, Martin R, Burrell LJ, Grabsch EA,
Kirsa SW, O’Keeffe J et al. Efficacy of an alcohol/
chlorhexidine hand hygiene program in a hospital with
high rates of nosocomial methicillin-resistant
Staphylococcus aureus (MRSA) infection. Med J Aust 2005;
183: 509–14.
Pronovost P, Needham D, Berenholtz S, Sinopoli D,
Chu H, Cosgrove S et al. An intervention to decrease
catheter-related bloodstream infections in the ICU. N
Engl J Med. 2006; 355: 2725–32.
Rybak MJ. Antimicrobial stewardship. Pharmacotherapy
2007; 27: 131S–35S.
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
27 Australian Commission on Safety and Quality in
Healthcare. Antibiotic stewardship program. 2008 [cited
2008 Nov 1]. Available from: URL: http://
www.safetyandquality.gov.au/internet/safety/
publishing.nsf/Content/PriorityProgram-03#five
28 Therapeutic Guidelines: Antibiotic, Edition 13. Melbourne,
VIC: Therapeutic Guidelines Limited; 2006.
29 Carman WF, Elder AG, Wallace LA, McAulay K,
Walker A, Murray GD et al. Effects of influenza
vaccination of health-care workers on mortality of
elderly people in long-term care: a randomised
controlled trial. Lancet 2000; 355: 93–7.
30 Andrews RM, Skull SA, Byrnes GB, Campbell DA,
Turner JL, McIntyre PB et al. Influenza and
pneumococcal vaccine coverage among a random sample
of hospitalised persons aged 65 years or more, Victoria.
Commun Dis Intell 2005; 29: 283–8.
Appendix 1
Ten commandments of infection prevention for
physicians
1. Always disinfect your hands with alcohol-based
hand rub BEFORE and AFTER touching a patient or
performing a procedure. Set the example for your team
and expect others to follow your lead.
2. Dress well for safer care – abandon ties and lanyards,
bare your arms to the elbow – no wrist watches or
jewellery.
3. Insist on the provision of alcohol-based hand rubs at
the patient bedside and in your clinic/rooms.
4. Take alcohol-impregnated wipes on your ward
rounds to disinfect equipment, such as stethoscopes and
pulse oximeters between use on every patient.
5. Ensure your team follows a standard, methodical,
sterile (aseptic) approach for all invasive procedures
(especially IV line insertion).
6. Invasive devices are potentially dangerous – remove
them as soon as you can (within 3 days for peripheral
cannulae).
7. Target antimicrobial therapy – consult Therapeutic
Guidelines: Antibiotic for the most appropriate agent(s),
dose, route and duration.
8. Be the first on your team to have the influenza
vaccine every year and make it known to others.
9. If you’re not receiving regular, relevant feedback
about healthcare-associated infections like MRSA involving your patients, then you’re missing out – insist on it.
10. Look beyond the obvious when seeking source(s) of
infection. Surgical wound and device-related infection
may be present even in the absence of visible local
inflammation.
581
Internal Medicine Journal 39 (2009) 582–587
O R I G I N A L A RT I C L E
Between a rock and a hard place: exploring the conflict
between respecting the privacy of patients and informing
their carers
imj_2020
1
582..587
2
M. Gold, J. Philip, S. McIver1 and P. A. Komesaroff3
1
Palliative Care Service, The Alfred Hospital, 2St Vincent’s Palliative Care Services and Centre for Palliative Care Education and Research, and 3The
Monash Centre for Ethics in Medicine and Society, Monash University, Clayton, Melbourne, Victoria, Australia
Key words
communication, carers, patient care.
Correspondence
Michelle Gold, Palliative Care Service, The
Alfred Hospital, PO Box 315, Prahran 3181,
Melbourne, Vic., Australia. Email: M.Gold2@
alfred.org.au
Received 9 November 2008; accepted 23 April
2009.
doi:10.1111/j.1445-5994.2009.02020.x
Abstract
Background: A patient’s right to privacy is considered fundamental to
medical care, with physicians assuming the role of guardian of the clinical
information which is conveyed to the patient. However, as a patient’s health
declines, physicians are often challenged by the need to protect patient privacy
while addressing the expectations of the patient’s carers, who seek medical
information to provide appropriate care at home.
Aims: This study sought to explore the expectations of patients, their carers
and physicians regarding the communication of clinical information to carers.
Methods: Surveys were distributed in outpatient clinics at a metropolitan
quaternary hospital, with responses from 102 patients and carers, as well as
219 medical staff.
Results: The expectations of patients and carers differed from those of medical
staff. Physicians typically believed discussions with carers should begin following the patient’s permission and at the patient’s request. Patients and carers,
however, believed information should be automatically offered or provided
when questioned. Further, carers generally felt information updates should
occur regularly and routinely, whereas physicians indicated updates should
occur with prompting either by a major clinical change or in response to a
carer’s concern.
Conclusion: Physicians should be aware that the expectations of patients and
carers regarding information communication to carers may not match their
own. Meanwhile, patients and carers should be made aware of the constraints
upon physicians and should be encouraged to convey their preferences for
information sharing. These tasks could be facilitated by the development of a
prompt sheet to assist the clinical encounter.
Introduction
Doctors are often criticized by family carers of their
patients for failing to provide adequate information. This
Funding: The authors gratefully acknowledge financial assistance from the Bethlehem Griffiths Research Foundation who
provided funding for this project.
Conflict of interest: None.
582
occurs in an era where the community and its legislators1
demand respect for the privacy of patient information
and expect that doctors will safeguard this. How are
doctors to negotiate a mutually satisfactory outcome to
this apparent conflict?
The philosophical tradition of Western medicine
emphasizes the role of the rational, autonomous patient,
who is assumed to be at the centre of all decision-making
processes. It is widely accepted that patients should be
able to choose how information about them is collected
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Conflict between patients’ privacy and carers’ information needs
and disseminated including, in particular, whether it is
shared with family and friends. However, this theoretical
model of patient-centred care does not always reflect the
clinical realities. In practice, there are several settings in
which access to information by family members or carers
becomes a clinical concern.
Modern healthcare systems increasingly require
complex forms of medical care to be undertaken in the
patient’s home. Accordingly, many patients with significant disabilities rely upon a carer to function successfully
in the community. Clear communication with the carer is
essential, as they often take on a quasi-clinical role.
However, when people with significant morbidity and
care needs are hospitalized, medical and health information is usually only disclosed to carers if explicit permission is given to professional staff to include others. This
can lead to the perplexing situation in which a relative
who has played a central role in the care of a patient and
is deeply concerned with his or her medical problems
may be denied information, or, at the least, receives little
communication from staff. In contrast to this situation,
where a patient is not competent to make decisions for
him or herself, such as in severe illness or dementia,
families are routinely called upon to participate directly
in decision making, for which full access to medical information is offered as a natural consequence.
These examples represent two divergent positions
taken in the communication relationships between physicians, patients and their carers. They indicate that the
belief in the primacy of the patient within the therapeutic
relationship can give rise to tensions which affect the
availability of health information. In addition, the needs
of patients’ carers to provide support and actively participate in decision making deserve recognition. The intensity of these tensions and their implications vary
according to the particular setting.
Although the perceived information needs of patients
have been extensively discussed in the biomedical
literature,2–5 those of the carers have received less attention. Morris and Thomas reported that the provision of
information to carers has many benefits, including mobilizing coping strategies, decreasing anxiety, developing
competency as a carer, and enhancing problem solving.6
The same researchers noted that in cases of cancer, carers
expect to be given access to information about patients’
illnesses.7 However, relatively little is known about the
expectations of carers more broadly, or about the expectations of patients and their physicians regarding the
sharing of medical information. We therefore set out to
examine the views of patients, carers and physicians, in
order to clarify the demands and expectations from all
parties regarding the sharing of medical information with
carers.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Methods
Study sample and procedures
For the purposes of this study, a ‘carer’ was defined as a
non-paid family member or friend who was identified by
a patient as being directly involved in his or her care.
Separate, parallel surveys were developed for (i) competent patients, (ii) their carers and (iii) medical staff. Translated versions of the surveys in Greek or Russian (the
languages other than English most commonly spoken at
the study institution) were provided for those who preferred them. Inclusion criteria included ages at least
18 years, fluency in written English, Greek or Russian
and capacity to complete the survey.
Surveys were piloted with small groups of subjects to
confirm content and face validity.
They were then distributed to patients and their carers
attending medical and surgical outpatient clinics at a quaternary teaching hospital. General Medicine, Oncology,
General Surgery and Urology clinics were chosen to
ensure a broad cross-section of diagnoses and clinical
disciplines. Consecutive patients and carers were
approached to participate in the study upon their arrival
at designated clinics. As such, participants were randomly
recruited, because it was unknown to the investigators
who would be attending which clinic and when.
The study was explained to all participants by the
research assistant, and information sheets and consent
forms were also provided. After providing verbal consent,
patients and carers completed the surveys independently
of each other. Participants could provide more than a
single response as appropriate. The study sought
responses from 50 matched patients and carers to provide
an adequate sample for a detailed exploration of the
subject under investigation. The physician surveys were
sent to all medical staff employed at the hospital, and
followed by two electronic reminders. No incentives to
participate were offered. The completion and return of
surveys signalled consent from the medical staff. All the
surveys were anonymous. The project was approved by
the institutional Human Research and Ethics Committee.
Results
Study sample
A total of 104 participants, comprising 52 patient and
carer pairs, was recruited from outpatient clinics and
included in the study. Two respondents who provided
incomplete data were excluded, resulting in a final data
set of 102 participants (51 patient and carer pairs),
representing a response rate of 98%. Specifically, these
583
Gold et al.
Table 1 How should carers obtain medical information about the
patient?
Responses
P%
C%
M%
Directly from the patient
From the doctor but only when the patient is present
From the doctor after the patient has given permission
From the doctor when the carer has a question
From the doctor without the patient’s knowledge
Families should not have access to the information
Other, please specify
42
30
40
52
6
8
2
28
46
36
50
4
0
8
43
29
83
13
1
0
2
P = patients (n = 51); C = carers (n = 51); M = medical staff (n = 219); more
than one option could be selected.
included General Medical (n = 20), Medical Oncology
(n = 30), General Surgery (n = 30) and Urology (n = 22)
services. Two hundred and nineteen responses were
received from medical staff following the targeted dissemination of 567 surveys, representing a response rate
of 38%.
Mean age for patients was 67 years (range 47–96 years,
SD = 16.56). For carers, the mean age was 55 years
(range 32–94 years, SD = 15.86). None of the respondents elected to complete the survey in Greek or Russian.
The physician cohort consisted of 137 men and 74
women including junior and senior medical staff from a
range of disciplines within the hospital.
Apparent differences in views among the three groups
were revealed in relation to three principal areas: (i) How
and what information should be given; (ii) When information should be given; and (iii) When updates of
information should occur.
How and what information should be given?
When asked how carers should obtain medical information about the patient, most patients (52%) and carers
(50%) indicated that the doctor should provide the information whenever a carer has a question (Table 1).
Further, 46% wanted the doctor to provide information
in the presence of the patient.
By contrast, 83% of medical staff reported that information should come from the doctor only when the
patient has given prior permission. While the majority of
patients (88%) and carers (84%) indicated that carers
should be told everything about a patient’s condition,
most doctors (53%) reported that carers should only
be given information already known to the patient
(Table 2).
Table 2 How much information should carers be given?
Responses
P%
C%
M%
Everything about the condition
Only basic information
Only information the patient is aware of
No information at all
88
2
14
0
84
4
8
0
22
5
53
24
P = patients (n = 51); C = carers (n = 51); M = medical staff (n = 219); more
than one option could be selected.
and provide information to them (Table 3). However, this
view was shared by only 20% of doctors in relation to
outpatient care and 26% of doctors in relation to inpatient care. The majority of doctors reported that information should be provided in response to a request from a
patient, with similar figures for outpatient (67%) and
inpatient (71%) care.
When should updates of information occur?
In the outpatient setting, most doctors (60%) reported
information to carers should be updated when the latter
asked directly or expressed a specific concern. By contrast, most patients and carers (70%) reported updates
should be given routinely with every outpatient visit
(Table 4). In the inpatient setting, doctors (73%) indicated carers should be updated if there were a major
change in the patient’s condition and 48% of patients
agreed with this view (Table 4). In contrast, most carers
(54%) indicated they should receive updated information from doctors after every inpatient consultation.
Written comments made by physicians highlighted an
awareness of problems created by mismatches of communication. One doctor commented:
I think doctors often find families demanding, defensive, protective, entitled, accusatory. Families, on the
other hand, feel disempowered, kept in the dark,
ignored, misunderstood. It is a ‘vicious circle’ situation.
Table 3 When should information be given to the carer?
Responses
P%
C%
M%
(outpatients)
M%
(inpatients)
80
72
20
26
When should information be given?
Doctors should do this
automatically
At the carer’s request
At the patient’s request
Other, please specify
18
10
0
10
20
2
22
67
11
24
71
9
Most patients (80%) and carers (72%) indicated that
doctors should automatically begin discussion with carers
P = patients (n = 51); C = carers (n = 51); M = medical staff (n = 219); more
than one option could be selected.
584
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Conflict between patients’ privacy and carers’ information needs
Table 4 How often should doctors update information for the carer (re: outpatients and inpatients)?
Responses
After every visit
When there is a major change
When carers ask
Not at all
Patients (%)
Carers (%)
Medical staff (%)
Out
In
Out
In
Out
In
70
34
26
0
24
48
32
0
70
34
22
0
54
34
26
0
15
51
60
5
4
73
47
0
Patients (n = 51); carers (n = 51); medical staff (n = 219); out = when patient is an outpatient; in = when patient is an inpatient; more than one option could
be selected.
A number of doctors expressed a belief that each situation was unique, involving its own set of complexities,
and suggested that decisions should be made on a caseby-case basis rather than based on a rigid formula. The
importance of such complexities was highlighted in comments about possible circumstances that might affect
decisions about whether patients and carers should be
given the same amount of information. Such circumstances often related to the disclosure of a poor prognosis:
Certainly there are many patients who do not want to
know all the details and equally carers who feel they
need to, to feel comfortable managing the situation –
very individual.
Physicians noted that in certain cultural traditions,
families often insist on limiting the information given to
patients, and expressed a variety of personal responses to
this situation. One doctor felt that it might be culturally
appropriate to comply with a family’s wishes, whereas
another described a similar situation where the withholding of information was refused. Sometimes the wish to
limit information came from the patient themselves. As
one doctor noted:
I have been in the situation of having explained to a
patient that they were terminally ill, and having them
express a wish NOT to discuss their prognosis further. I
have respected this wish, and (with the patient’s
express permission) honestly answered ongoing questions from their family.
A number of physicians spontaneously suggested strategies to improve communication with carers. These
included three main approaches: (i) encouragement of
carers to contact the medical team and pro-actively ask
questions; (ii) making carers aware of policies restricting
the dissemination of information without a patient’s
permission; and (iii) nomination of a particular
family spokesperson to avoid the need for multiple
explanations.
While patients and carers were invited to submit additional comments these were generally very brief. A small
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
number of patients and carers did express a desire for the
patient not to be made aware of a terminal diagnosis.
Discussion
The views of patients and carers regarding the dissemination of information to carers were generally similar.
There was, however, some disparity between their views
and those of physicians. Physicians indicated carers
should be given information only after the patient had
given permission, or at the patient’s request. They did not
see their role as updating information routinely, but
would do so in response to a specific prompt, such as a
change in the condition of the patient or a carer’s
concern. This was in contrast to carers and patients who
believed the carer should be given information automatically at each consultation.
These alternative perspectives will not surprise clinicians. Indeed, many physicians in this study commented
upon the mismatches that occur in everyday care and
communication with carers. The possible sources of these
different expectations deserve consideration. One explanation resides in the nature of relationships. The carers of
patients are frequently spouses and lifelong partners. In
the event of one partner becoming ill, there is an expectation that the information flow will continue unchanged
within their relationship. Patients and their carers often
tend to view themselves as a unit. Equally, doctors must
remain mindful of the possibility that some patients do
not wish to share their medical information with family
members, for a variety of reasons.
For physicians, an alternative set of imperatives typically operate. For example, physicians often view the
patient as the centre of care, with the patient’s primacy in
the relationship and their right to privacy, enshrined in
healthcare legislation. Seemingly, patients and carers are
formulating their expectations in a climate of open relationship, while physicians are being guided by a climate
of ethical and legislative limitations. The sometimes contradictory nature of these approaches impacts upon the
communication with patients’ carers.
585
Gold et al.
A second possible explanation may lie in the relative
responsibility assigned to the task of communicating
information. Carers reported they should receive updated
information from the doctor after every consultation,
regardless of whether the patient was in or out of hospital. However, physicians indicated they are more likely to
wait for a trigger to prompt information sharing with
carers, such as a concern or a major change, and then
only after the patient has given permission. Indeed, the
workload carried by physicians is likely to influence the
priority assigned to informing carers. Physicians are typically involved in the medical care of many patients, and
the sharing of information with many carers. Time constraints suggest that some prioritization must occur, and it
appears that physicians give priority to spontaneous
information sharing at times of major clinical change. For
carers, the patient is the main and only priority, and
therefore, automatic updates are expected. Given this
discrepancy, carers might believe they are insufficiently
informed, possibly resulting in dissatisfaction that could
potentially be prevented.
A slight divergence was apparent between responses
from patients and carers when asked how often physicians should update information for the carer in the case
of patient hospitalization. In contrast to the majority of
carers who preferred frequent updates, the highest percentage of patients indicated updating of information
should occur with major change, perhaps suggesting
an awareness of the physicians’ workloads and time
constraints.
There are limitations to this study which deserve
comment. The response rate of physicians (38%) limits
the ability to generalize findings to all hospital physicians.
However, studies surveying physicians have reported
response rates varying between 17.6%8 and 56%,9 suggesting 38% may be an acceptable response rate for
studies involving time-constrained physicians. The total
number of respondents (219) represents a large cohort
whose views should not be disregarded. Further, results
from the present study support findings from several
comparable studies examining the views of patients,
carers and healthcare professionals which have also
identified the problematic nature of mismatched
communication.2–4 The lack of patient and carer respondents who only spoke Greek or Russian prevented a
comparative cross-cultural analysis.
Notwithstanding these limitations, results from this
exploratory study have identified potentially useful findings. Comments made by responding physicians suggested that the research question under examination
resonated strongly with them, indicating potential clinical relevance of the study. This relevance was reinforced
by the unprompted, possible solutions suggested by
586
physicians to resolve the perceived mismatch of communication. Physicians appeared mindful of the problems,
and were apparently interested in strategies that may
circumvent the disparate expectations.
This study raised some key questions about the nature
of the relationship between patients and physicians. As
previously noted, physicians typically see the patient
as the sovereign of his or her medical information.
However, in the example of patient incompetency, physicians often reveal information to family members
without hesitation. Similarly, where a patient is suffering
from a terminal illness and is in need of palliative services, the family is often automatically included in the
unit of care. In this instance, the restriction of information to the patient is subject to the approach of palliative
care which includes both patient and his or her family as
the unit of care.10
It is possible that as patients move through a
continuum of an illness, when communicating with
physicians, the role of themselves and of their carers
continuously changes. At some point, the patient may be
regarded by their physician as controlling his medical
information, but with declining health, the role of the
carer increases in importance. It is unclear if patients are
aware that their role and that of their carer changes as
illness increases. Nor is it clear that physicians are cognisant of their changing regard for the role of carers in this
communication relationship. Future studies should
investigate in detail these questions of communication
between patients, carers and physicians in an attempt to
characterize and elaborate the models of care in medicine, and at different stages of illness.
The results from this study identified possible disparities between the expectations of patients, carers and physicians, when considering the sharing of information.
Physicians should be aware that patient and carer expectations may not be uniform and may differ from their
own. Meanwhile, patients and carers should be aware of
the importance of conveying their preferences for information sharing, as well as the constraints that physicians
work within.
These findings support a case for developing an educational prompt sheet for patients and carers offering
information and guidance regarding the provision of
clinical information to carers. Such a prompt sheet may
provide information regarding the usual practice of physicians in providing information to carers, as well as suggestions to guide patients and carers in alerting their
physician about particular concerns. Successful implementation of such a prompt sheet into routine clinical
care may lead to greater satisfaction with the healthcare
system for patients and carers, as well as improved
healthcare outcomes.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Conflict between patients’ privacy and carers’ information needs
Future research should be directed towards characterizing the nature of the communication relationship
between the patient, the carer and the physician at different points in the illness trajectory.
References
1 Breen K, Pluekhan V, Cordner S. Ethics, Law and Medical
Practice. Sydney: Allen & Unwin; 1997.
2 Clayton JM, Butow PN, Tattersall HN. When and how to
initiate discussion about prognosis and end-of-life issues
with terminally ill patients. J Pain Symptom Manage 2005;
30: 132–44.
3 Exley C, Field D, Jones L, Stokes T. Palliative care in the
community for cancer and end-stage cardio-respiratory
disease: the views of patients, lay-carers and health care
professionals. Palliat Med 2005; 19: 76–83.
4 Kirk P, Kirk I, Kristjanson LJ. What do patients receiving
palliative care for cancer and their families want to be
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
5
6
7
8
9
10
told? A Canadian and Australian qualitative study. BMJ
2004; 328: 1343–7.
Stewart MA. Effective physician-patient communication
and health outcomes: a review. Can Med Assoc J 1995;
152: 1423–33.
Morris S, Thomas C. The need to know; informal carers
and information. Eur J Cancer 2002; 11: 183–7.
Morris S, Thomas C. The carer’s place in the cancer
situation: where does the carer stand in the medical
setting? Eur J Cancer 2001; 10: 87–95.
Chao C. Physicians’ attitudes toward DNR of terminally
ill cancer patients in Taiwan. J Nurs Res 2002; 10:
161–7.
Mulcahy P, Buetow S, Osman L et al. GPs’ attitudes to
discussing prognosis in severe COPD: an Aukland (NZ) to
London (UK) comparison. Fam Pract 2005; 22: 538–40.
Philip J. Lowering One’s Net Deeper and Deeper: the
cultural and ethical components of palliative care. [PhD].
Melbourne, Monash; 2007.
587
Internal Medicine Journal 39 (2009) 588–594
O R I G I N A L A RT I C L E
Understanding organ donation in the collaborative era:
a qualitative study of staff and family experiences
imj_1826
588..594
S. L. Thomas, S. Milnes and P. A. Komesaroff
Centre for Ethics in Medicine and Society, Department of Medicine, Monash University, Melbourne, Victoria, Australia
Key words
organ donation, intensive care unit staff
experience, National Organ Donation
Collaborative, family experience, qualitative
study.
Correspondence
Samantha L. Thomas, Centre for Ethics in
Medicine and Society, Department of
Medicine, Monash University, Commercial
Road, Prahran, Vic. 3181, Australia.
Email: [email protected].
edu.au
Received 7 June 2008; accepted 31 July 2008.
doi:10.1111/j.1445-5994.2008.01826.x
Abstract
Background: Despite the success of the Breakthrough Collaborative
Methodology (BCM) in increasing organ donation rates there has been little
published evidence on the effect of the BCM on the wider attitudes and
experiences of those involved in organ donation. This study sought to identify
whether the National Organ Donation Collaborative in Australia had any
additional influence on improving the experiences of staff and family members
in the organ donation process.
Methods: In-depth qualitative interviews with 17 family members from 13
families who had agreed to the organ donation of a deceased relative and 25
nurses and intensive care specialists at the Alfred Hospital, Melbourne,
Victoria were carried out.
Results: The key factor in family members’ decision to donate was prior
knowledge of the deceased’s donation wish. Although most family members
did not regret their decision to donate, many were deeply dissatistified and, at
times, confused by the technical and administrative nature of the donation
process. Most staff members commented that the key community message
about donation should be to encourage people to discuss donation rather than
urging people to sign donor registers.
Conclusion: This study identified valuable insights into the processes by
which family members and intensive care unit staff deal with the actual
processes of donation. Findings suggest that the process for families is far more
complex than a simple agreement or refusal to donate. This study suggests that
we should not assume that ‘rates’ of donation in Australia would increase
merely through administrative programmes or marketing campaigns.
Introduction
In policy and media campaigns aimed at increasing donation rates in developed societies’ organ donation is conceptualized as an altruistic act and framed as a ‘gift of life’.1
Many factors may, however, influence a family’s decision
to donate or refuse organ donation at the end of life of a
loved one. Some suggest that there is a tension between,
on the one hand, the technical and administrative pro-
Funding: Australians Donate.
Conflict of interest: Paul Komesaroff was a member of the Board
of Australians Donate.
588
cesses and the need to increase organ donation rates to
meet the demand of those waiting for transplants, and, in
contrast, the complex and layered nature of death, including the cultural processes and complexities associated with
death, grief and decision-making at the end of life.2 Therefore, organ donation rates may not merely reflect an
administrative issue – to be rectified by policies to encourage donation through incentives, education, donor cards
or ‘opt-out’ systems – but may in fact show deeper underlying processes in which there may be a natural limit to
organ donation rates in a multicultural society.
The Breakthrough Collaborative Methodology (BCM)
has been shown to be highly successful in the USA,
with consistent increases in the rates of organ donation
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Staff and family experiences of organ donation
between 2003 and 2005.3,4 This collaborative process has
been well documented elsewhere.5,6 In brief, however,
the BCM is based on organizations working together,
identifying best practices and developing clear strategies
to achieve common goals and targets. Given the success
in increasing organ donation rates using the BCM in the
USA, and given that Australia was noted to have one of
the lowest rates of organ donation in the world, the
National Organ Donation Collaborative (NODC) was
implemented in 22 hospitals across Australia in July
2006. The NODC is an ongoing initiative and has already
been shown to have a significant influence on rates of
organ donation as compared with previous years.
Despite the apparent success of the BCM on organ
donation rates there has been little published evidence on
the effect of the BCM on the wider attitudes, satisfaction
and experiences of those involved in procuring organs and
for the families who may agree or not agree to organ
donation in deceased relatives. There is also conflicting
evidence about the effect of organ donation and the organ
donation process on staff and family members. For
example, studies have suggested that organizational
factors within hospitals may influence organ donation
rates, such as (i) strained relationships with surgical teams
may negatively influence nurses’ stress levels,7 (ii) nurses
are more positive about organ donation when they are less
involved with the surgical process of organ donation or
retrieval,8 (iii) the reluctance of nursing and medical staff
to address issues of death and organ procurement with
family members suggests that physicians may experience
a patient’s death as a personal failure that they are reluctant to share with others, (iv) minority families may be
given less opportunity to discuss donation as compared
with white families,9 (v) and the timing of the donation
discussion, information and context of the discussion,
attitudes and beliefs, and overall satisfaction of the healthcare team may influence a donation decision.10,11
Whereas a core aim of the NODC is to ‘bring together
a group of health organisations which are committed to
redesigning and improving their systems to achieve a
major, rapid improvement in the quality and safety of
healthcare; in this case, an increase in organ donation
rates’, this study sought to identify whether the NODC in
Australia has had any additional influence on improving
the experiences of staff and family members in the organ
donation process.12
Methods
The choice of a qualitative method
We used a descriptive qualitative method to enable us to
gather rich narratives about the experiences of people
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
involved in organ donation and the relationships
between health professionals, family members and the
healthcare system. An interview schedule was based on
themes identified through extensive consultation with a
wide variety of stakeholders involved in Organ Donation,
including family members, intensive care unit (ICU) physicians and nurses, organ donation coordinators (ODC)
and policy-makers.
Source population
The study was conducted in the ICU at The Alfred Hospital in Victoria. The Alfred is an important tertiary referral teaching hospital in Australia, with a major role in the
provision of specialist services both throughout the state
of Victoria and nationally. As a National Trauma Centre,
The Alfred was chosen because it has one of the highest
ongoing rates of organ donation in Australia and was one
of the 22 hospitals involved in the NODC.
Sampling strategy
The study focused on family members who had agreed to
organ donation in relatives who had been declared brain
dead and on staff members (ICU physicians and nurses)
involved in organ donation in the 12 months before (July
2005 to June 2006) and after (July 2006 to June 2007)
the implementation of the NODC.
Recruitment
In accordance with ethics committee requirements a list
of family members involved in organ donation was provided by Lifegift the Victorian Organ Donation Service.
The Organ Donation Coordinator (ODC) involved in
working with each family made first contact with the
family and sought consent for family members to be
contacted by the research team. A member of the
research team then provided further written and verbal
information about involvement in the study and family
members were given 1 week to decide to whether to
participate in the research.
Interviews with family members
Semistructured, face-to-face interviews were conducted
to examine the study aims with participants. The
interviews took place between 2 and 18 months postbereavement. The dates of the interviews were
specifically organized so as to avoid coinciding with any
significant family events or anniversaries, such as the
date of the donor’s death, the donor’s birthday or family
holidays, such as Christmas. All interviews were
589
Thomas et al.
conducted at the homes of family members. Interviews
took between 1 and 3 hours to complete.
Interviews with hospital staff
Semistructured, face-to-face and telephone interviews
were conducted to learn about staff experiences of donation at The Alfred Hospital. Interviews were conducted
with ICU physicians and nurses. Interviews took between
30 and 60 min to complete.
Data analysis
Data analysis was based on rigorous qualitative techniques that developed analytical categories, tested our
processes of analysis and then provided an explanation of
why these categories occurred. A constant, continuous,
comparative method of analysis, guided by grounded
theory (the inductive process of identifying analytical
categories as they emerge from the data), was used
throughout the study to investigate themes both within
and across groups, and to generate novel theories and
hypothesis.13 QSR NVIVO was used as a management
tool to help group the data into categories and theoretical
themes, although most of the analysis was conducted by
hand.
Results
Seventeen families were approached to take part in the
study – 7 who had agreed for donation between July
2005 and June 2006 – and 10 who had agreed to donation between July 2006 and June 2007. Four families
who were approached did not agree to take part in the
study. A total of 17 family members took part in the
study, including 7 mothers, 3 fathers, 4 wives, 2 sons and
1 sister. A total of 25 staff members took part in the study,
including 9 nursing staff, 11 intensive care specialists and
5 ODC. In our analysis, and unless otherwise specified,
we have use the terms ‘a few’ to refer to less than 25% of
all participants; ‘some’ to refer to between 25 and 50%
of participants; ‘many’ to refer to between 50 and 75% of
participants and ‘most’ to refer to more than 75% of
participants.
Family experiences
There were no differences in the quality of family experiences in those who agreed to donation either before or
after the introduction of the BCM. Although each family
had unique experiences, there were several common
themes that emerged across the two groups.
590
Satisfaction with hospital care
Some family members struggled with their initial
experiences when they first arrived at the emergency
department (ED). In particular, a few family members
expressed dissatisfaction with the lack of information that
was provided to them by ED staff about their relative.
Family members stated that they felt ‘isolated’, ‘lost’ ‘in
limbo’, ‘disappointed because they left me up in the air’,
that they ‘weren’t kept up to date’, or that it ‘seemed to
take forever’ until they found out about the condition of
their relative. Two family members commented that they
would have liked to have seen their relative in the ED
‘regardless of the state he was in’ and even if there ‘were
one hundred people working on him’. This may stress the
need for additional care and support of those with relatives in a critical condition on first arrival at hospital.
Most family members were happy with the care
received from the ICU team and donor coordinators
stating that they were ‘wonderful’, or that ‘you couldn’t
fault them’ or that they had ‘nothing but respect and
admiration’ for staff members. Family members spoke of
the additional care and support provided in particular by
ICU nurses. Families felt that they had received individual treatment and that they were not treated like just
another case. In particular, some family members spoke
of the emotional connection with ICU staff. It was
common for family members to discuss how much they
appreciated the openness of staff members. For example,
‘one of the nurses and I had a good cry together’ and
‘[the doctor] was a bit teary when he told us’.
While families were critical of the bureaucratic nature
of the donation process, most were very happy with the
support, care and information provided by donor coordinators and staff during the process.
Information and decision-making about donation
Most family members were happy with the explanation
they received about organ donation from ICU staff
members before making their decision to donate.
However, most prefaced this statement by saying that
they knew their relatives’ wishes to donate and that
they were very clear about these wishes when the
information was being given. Interestingly, although
most family members stated that they had not ‘consciously’ had an in-depth discussion about organ donation with their relative, they could, nonetheless, recall
at some point having at least a brief or ‘flippant’ conversation about donation.
We talked about it years ago when the kids were little.
I don’t know why. Somewhere along the line it came
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Staff and family experiences of organ donation
up. If anything happens to us or to them that organ
donation was what we thought would be the right
thing to do.
Only one family was unaware of their relative’s wishes
and stated that the decision to donate was a difficult one.
Although they appreciated the information provided by
hospital staff, they were very clear that this had no
bearing on their decision to donate. Rather, the decision
to donate was reached after long discussions within the
family. As such, there was no clear indication that the
hospital environment or the quality of care received had
an influence on any of the decisions to donate.
The donation process
Although family members were generally satisfied with
the lead up to their decisions about donation many
struggled with some of aspects of the donation process.
First, family members were surprised and at times negative about the length of the process. Many stated that
they did not realize that the ‘act’ of donating organs
would take such a long time Many also indicated that the
length of time it took to donate was both unexpected and
stressful, with one family stating ‘we had to sit there and
sit there and sit there and sit there and just watch and
look’. Although in hindsight individuals did not regret
their decision to donate they were critical of the donation
process with one person claiming that
The whole process was 10 hours of hell. Have the
organs but don’t make it so traumatic.
Many family members also commented that they were
unprepared for the highly technical information surrounding the process of donation, with one mother
describing the information as going ‘in one ear and out
the other’. A few described the process of the ‘nitty gritty’
of filling in the forms as particularly difficult with one
person describing it as ‘horrendous’. Many participants
referred to the bureaucracy associated with the process of
donation:
Paperwork. Sitting there in the office, family details
and all sorts of histories and authorities. It was just
absurd. The process is wrong. They’ve got to change it.
It’s not all about them getting their organs. It’s also
about the family dealing with the bigger picture and
the life long change.
For others, seeing their relative in the trauma unit
‘hooked up to all these machines’ or ‘watching the
oxygen’ was particularly confronting. Others could recall
with great detail the moments when their relatives had
their drips changed or giving their relative a bed bath.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Some family members reflected that organ donation was
a technical process outside the realms of what usually
happens when someone dies and was unexpected and
stressful.
After donation
Family members had many different experiences following donation. For some the ability to fulfil their relative’s
wishes seemed to bring some comfort when it was the
only sense they could make of a tragic situation.
I think it helped me. It helped me to know that he was
going to live on in somebody else. That helped me
tremendously.
However, many family members felt unsupported after
donation. Although most family members did not wish to
take part in formal counselling, they struggled to find
appropriate support mechanisms to help them. Families
dealt with the experiences of the death differently and
used a number of different coping strategies. One mother
spoke of ‘therapeutic’ importance of taking a plaster cast
of her son’s hands and feet with the ODC. Three family
members stated that the fact that the ODC – whom many
had formed very emotional bonds with and described as
their ‘advocate’ – had been present during the operation
to remove the organs was extremely important in helping
them feel that someone was ‘looking after’ their relative
during the donation. Others spoke of the difficulties they
experienced in ‘letting go of the detail’ of the experience,
or becoming preoccupied with not knowing where their
relative’s organs had gone, and the ‘cloak of secrecy’
surrounding recipients. Some family members spoke
about the anonymous letters that they had received from
the organ recipient, but were emotionally unable to write
a letter of reply. Others spoke of the desire to become
involved in speaking to the media, fund-raising or being
involved in promoting organ donation, so that they could
feel that ‘some good’ had come out of their own personal
tragedy. For others, wearing a red wristband from an
organ donation foundation gave continued meaning to
their experience.
There was some indication that the highly technical
and medicalized way in which death occurred meant that
family members required support for some time after the
donation had occurred, but could not access the support
they thought they needed. However, some family
members also could not say what exactly this support
would be and whether formalized support would actually
be useful or helpful. One father described looking for the
right kind of support as ‘banging your head up against a
brick wall’.
591
Thomas et al.
Staff experiences
The donation process
Nurses and physicians described the process of asking for
organs as a difficult and at times stressful experience.
Many stated that the topic of organ donation was at times
so ‘difficult to bring it up’ and ‘uncomfortable’ that they
did not always bring up the topic of donation because
‘you perceive a grieving family’. Some physicians commented that it was a ‘nerve wrecking experience’ and
was particularly difficult for junior staff members, who
were often ‘scared’ or felt ‘guilty for raising the subject’
rather than a ‘duty’ and ‘obligation to find out the wishes
of the person’.
Physicians and nursing staff spoke openly about the
wide range of reactions that families had when asked to
donate. These ranged from an ‘immediate yes’ to ‘shock
and horror that you could have asked’. Many nurses and
physicians spoke about the deep conflict that sometimes
existed within families over agreeing to donate.
I think the most tragic are where the majority wish for
the donation to take place, but there are one or two
who say ‘no I couldn’t let his body be cut up’. You see
the room suddenly start to agonise themselves about
what the right answer is.
able for discussions with the family, and for the family to
have time to think and discuss decisions with each other
and with staff, (iii) only discussing donation after brain
death and ensuring that the patients’ well-being is all
they are thinking about until brain death, (iv) an appropriate time gap between and acceptance of death and the
discussion about donation, and (v) that organ donation
was raised by staff who were specialized and trained in
having the discussion about donation.
Nurses and physicians spoke about the difficulties that
families had in understanding the concept of ‘brain death’.
Physicians commented that ‘no matter how well you tried
to explain brain death to families, you could almost guarantee that there would be one person in the room that
would not understand the concept’. However, many went
on to comment that despite this, most families did not
question that their relative was dead. One physician commented that family members ‘hear that word dead and
that is enough for them’. Nurses spoke about the confusion that family members experienced when they had
been informed that their relative was brain dead, but still
‘looked normal in many ways’. Four nurses stated that
family members were often surprised that nursing staff
still spoke to their relative when providing care, although
they had been told that they were ‘dead’.
Nurses spoke about the important role they played in
‘preparing’ families for the discussion about donation
with physicians, ‘recognising their grief’ and then in
‘consolidating’ ‘reassuring’ and ‘supporting’ families after
the discussion had taken place.
Most of the nurses and physicians agreed that the
waiting time between the declaration of death and the
harvesting of organs could be very stressful on family
members. Physicians in particular spoke of the need to be
extremely up front about the time that organ donation
may take, so that family members were not surprised or
distressed by the waiting involved.
General issues associated with the ODC
Recommendations about how to approach
donation
Some physicians felt that although the NODC had little
influence on the way in which people were asked about
donation at The Alfred, it had, however, promoted a
greater awareness and understanding about donation
among staff. Physicians also thought that a key goal of the
NODC could be to enhance collaborative working relationships between Emergency and Intensive Care staff to
help with the identification of potential ‘donors’ within
the ED.
Physicians took a variety of approaches in raising the
subject of organ donation with family members. For
example, some preferred to talk to only immediate family
members, whereas others stated that it was best to have
the discussion with the larger group of family and friends.
However, there were several common themes that arose:
(i) taking time with the family: physicians agreed that
talking through the issues slowly, not rushing the decision and ensuring that families fully understood what
they were agreeing to – particularly in terms of the
waiting time, (ii) ensuring that a private space was avail-
592
Many physicians and nursing staff commented that the
key community message about donation should be to
encourage people to discuss donation rather than exhorting individuals to donate organs. Some commented that
the process of asking if families agreed to donate would
be made easier if more families had had a discussion
about donation.
If they haven’t discussed it before the reaction varies
from initial confusion, don’t quite understand the
questions you know, question, what is organ donation? To, you must be joking.
Discussion
Although this study set out to determine the influence of
the Organ Donation Collaborative on family decisions to
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Staff and family experiences of organ donation
donate, it identified valuable information about how
family members and ICU staff deal with the actual processes of donation. It is important to acknowledge at this
juncture that this study was only conducted at The
Alfred; therefore the results may not be generalizable to
the experiences of family and staff members at other
hospitals.
Unlike other studies, negative experiences within the
hospital did not affect family members’ donation decisions. For example, although many had a negative experience within the ED they still agreed to donate. As has
been shown by other studies, the key factor in family
members’ decision to donate was knowledge of the
deceased’s donation wish.14 In many cases, family
members were able to recall discussions with the
deceased about donation, often sparked by a news item
or television show. Although these conversations were at
times only brief or fleeting, family members still based
their decision to donate on these discussions. This may
emphasize at a deeper level, the cultural factors that may
influence decisions to donate. As Burroughs et al. suggest
the signing of a donor card may not be enough to influence family decisions.15 Rather, remembering personal
discussions about the wishes to donate may be a critical
factor in donation decisions – giving family members the
opportunity to discuss beliefs and giving the family assurance and confidence in later decisions. Studies of end-oflife decisions show similar trends. For example, a study of
family involvement in end-of-life decision-making in ICU
found that the presence of a formal Advanced Directive
made no difference to the choices that family members
made when compared with those without Advanced
Directives.16 Rather, prior discussions with family
members and extensive communication with nurses and
physicians were key factors influencing family decisionmaking. Results indicate that although making the decision to donate was not easy for any family member,
discussions with the deceased and the support of other
family members in making the decision were important
factors. Yet, research has shown that many individuals do
not discuss their end-of-life wishes with family members.17 Barriers to discussions are predominantly sociocultural and include fear of death, trust in others to make
decisions, family dynamics and uncertainty about preferences. Thus, public health messages suggesting ‘having a
discussion’ with family members about donation may be
far more valuable in the long run than encouraging
individuals to sign a donor register or hold a donor
card.
Most family members stated that if they had the opportunity again they would make the same decision to
donate. However, most were deeply dissatistified and at
times confused by the technical and administrative
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
nature of the donation process. This was a troubling
finding and showed a conflict between the natural grieving process of the family and the way in which death
may become a ‘technical event’ when families agree to
donate. As in other studies we found that the family
members went through several complex stages in agreeing to donate.18 These included the initial shock of an
anticipated or sudden illness of a relative; the journey to
and arrival at the ED where they often could not see a
relative for a considerable amount of time; the transfer to
the ICU; the eventual diagnosis of ‘brain death’; the
request for organ donation; the process of deciding
whether to donate; the bureaucratic and legal processes
associated with donation; waiting for the harvesting of
organs; the final release of the body; and the short-term
and long-term grieving processes after death. Most ICU
staff members also agreed that there was a tension
between the technical nature of death associated with
organ donation and the need for families to say ‘goodbye’
to their relative. Most stressed the importance of allowing
families time to understand the information given, care
in the way and context that information was shared and
attention to their emotional needs. Most staff members
agreed that it was extremely important for families to
understand that the process of donation may be a lengthy
one.
Conclusions
Although much research has focused on the factors that
may ‘enhance or inhibit’ family agreement to organ
donation, this study has shown that the process for families is far more complex than simply agreeing or refusing
to donate.19 Furthermore, the findings indicate that we
should not assume that ‘rates’ of donation in Australia
would increase merely through administrative programmes or marketing campaigns. Although the key
factor in influencing a family’s decision to donate was
remembering a previous discussion with the deceased
relative about donation, this did not necessarily make
the ‘act’ of donation a positive experience for family
members. Further research is needed into (i) how families can be better supported during the donation process
and (ii) the conflict between the natural grieving process
and perhaps unintended long-term consequences of the
technical and administrative processes associated with
donation.
Acknowledgements
The authors gratefully acknowledge the support provided
by Australians Donate for the conduct of this study. Australians Donate was also the initiator and manager of the
593
Thomas et al.
National Organ Donation Collaborative. We are thankful
to The Alfred Hospital Intensive Care staff, in particular,
Dr Andrew Davies and Shena Graham and the ODC team
for their help in recruitment, and in discussing the results
of this study. Most importantly, we would like to thank
the families for taking the time to share their experiences
with us.
References
1 Wright L. Is presumed consent the answer to organ
donation shortages? No. BMJ 2007; 334: 1089.
2 Sque M, Long T, Payne S, Allardyce D. Why relatives do
not donate organs for transplants: ‘sacrifice’ or ‘gift of
life’? J Adv Nurs 2008; 61: 134–44.
3 Punch JD, Hayes DH, LaPorte FB, McBride V, Seely MS.
Organ donation and utilization in the United States,
1996–2005. Am J Transplant 2007; 7(5 Pt 2):
1327–38.
4 Marks WH, Wagner D, Pearson TC, Orlowski JP,
Nelson PW, McGowan JJ et al. Organ donation and
utilization, 1995–2004: entering the collaborative era.
Am J Transplant 2006; 6(5 Pt 2): 1101–10.
5 The Breakthrough Series: IHI’s Collaborative Model for
Achieving Breakthrough Improvement. IHI Innovation Series
White Paper. Boston: Institute for Healthcare
Improvement; 2003.
6 Mathew TH, Chapman JR. Organ donation: a chance
for Australia to do better. Med J Aust 2006; 185:
245–6.
7 Regehr C, Kjerulf M, Popova SR, Baker AJ. Trauma and
tribulation: the experiences and attitudes of operating
room nurses working with organ donors. J Clin Nurs
2004; 13: 430–37.
8 Cantwell M, Clifford C. English nursing and medical
students’ attitudes towards organ donation. J Adv Nurs
2000; 32: 961–8.
594
9 Siminoff LA, Mercer MB, Arnold R. Families’
understanding of brain death. Prog Transplant 2003; 13:
218–24.
10 Rodrigue JR, Cornell DL, Howard RJ. Organ donation
decision: comparison of donor and nondonor families.
Am J Transplant 2006; 6: 190–98.
11 Morgan SE, Stephenson MT, Harrison TR, Afifi WA,
Long SD. Facts versus ‘Feelings’: how rational is the
decision to become an organ donor? J Health Psychol
2008; 13: 644–58.
12 http://www.organdonation.org.au/
12799+0+national-collaborative.htm
13 Glaser BG, Strauss AL. The Discovery of Grounded Theory.
Strategies for Qualitative Research. New York: Aldine
Publishing Company; 1967.
14 Sque M, Long T Payne S. Organ donation: key factors
influencing families’ decision-making. Transplant Proc
2005; 37: 543–6.
15 Burroughs TE, Hong BA, Kappel DF, Freedman BK. The
stability of family decisions to consent or refuse organ
donation: would you do it again? Psychosom Med 1998;
60: 156–62.
16 Bernal EW, Marco CA, Parkins S, Buderer N, Thum SD.
End-of-life decisions: family views on advance directives
(Epub 2007 Jun). Am J Hosp Palliat Care 2007; 24:
300–307.
17 Glass AP, Nahapetyan L. Discussions by elders and adult
children about end-of-life preparation and preferences.
Prev Chronic Dis 2008 [cited 2008 Oct]: 5(1). Available
from URL: http://www.cdc.gov/pcd/issues/2008/jan/
07_0141.htm
18 Kesselring A, Kainz M, Kiss A. Traumatic memories of
relatives regarding brain death, request for organ
donation and interactions with professionals in the ICU.
Am J Transplant 2007; 7: 211–17.
19 Siminoff LA, Arnold RM, Hewlett J. The process of organ
donation and its effect on consent. Clin Transplant 2001;
15: 39–47.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Internal Medicine Journal 39 (2009) 595–599
O R I G I N A L A RT I C L E
What we have here is a failure to communicate! Improving
communication between tertiary to primary care for chronic
heart failure patients
imj_1820
1,2,3
S. Shakib,
1
4
595..599
1
H. Philpott and R. Clark1,4
Department of Clinical Pharmacology and Drug Optimisation Clinic, 2National Institute of Clinical Studies and 3Royal Adelaide Hospital,
Faculty of Health Sciences, University of South Australia, Adelaide, South Australia, Australia
Key words
chronic heart failure, discharge management,
primary care, continuum of care.
Correspondence
Sepehr Shakib, Clinical Pharmacology and
Drug Optimisation Clinic, Royal Adelaide
Hospital, Level 7, Emergency Block, North
Terrace, Adelaide, South Australia, Australia.
Email: [email protected]
Received 22 February 2008; accepted
22 August 2008.
doi:10.1111/j.1445-5994.2008.01820.x
Abstract
Background: The aims of this study were to determine the documentation of
pharmacotherapy optimization goals in the discharge letters of patients with
the principal diagnosis of chronic heart failure.
Methods: A retrospective practice audit of 212 patients discharged to the care
of their local general practitioner from general medical units of a large tertiary
hospital. Details of recommendations regarding ongoing pharmacological and
non-pharmacological management were reviewed. The doses of medications
on discharge were noted and whether they met current guidelines recommending titration of angiotensin-converting enzyme inhibitors and betablockers. Ongoing arrangements for specialist follow up were also reviewed.
Results: The mean age of patients whose letters were reviewed was 78.4 years
(standard deviation ⫾ 8.6); 50% were men. Patients had an overall median of
six comorbidities and eight regular medications on discharge. Mean length of
stay for each admission was 6 days. Discharge letters were posted a median
of 4 days after discharge, with 25% not posted at 10 days. No discharge letter
was sent in 9.4% (20) of the cases. Only six (2.8%) letters had any recommendations regarding future titration of angiotensin-converting enzyme
inhibitors and 6.6% (14) for beta-blockers. Recommendations for future
non-pharmacological management, for example, diuretic action plans, regular
weight monitoring and exercise plans were not found in the letters in this audit.
Conclusion: Hospital discharge is an opportunity to communicate management plans for treatment optimization effectively, and while this opportunity
is spurned, implementation gaps in the management of cardiac failure will
probably remain.
Introduction
Chronic heart failure (CHF), which affects more than
300 000 Australians, is a common and serious condition,
with a 5-year mortality rate of 50%.1,2 Fortunately, over the
Funding: Dr Sepehr Shakib is a Fellow of the National Institute
of Clinical Studies (NICS) is supported by the South Australian
Department of Health. Dr Robyn Clark is also a former NICS
scholar.
Conflict of interest: None
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
last 10 years, the prognosis has improved and clear survival
benefit has been shown with the use of angiotensinconverting enzyme inhibitors (ACEI), beta-blockers and
non-pharmacological therapies.3 However, there continue
to be significant gaps between best and actual practice, both
in hospital and community settings.3 CHF patients have
a significant rate of readmission, which can be lessened with
adequate post-discharge management.4
A previous audit by our group has found that more than
90% of eligible patients with CHF were prescribed ACE
inhibitors on discharge.5 In contrast, there were several
595
Shakib et al.
system barriers to the up-titration ACE inhibitors and the
addition of beta-blocker therapy identified during the
admission.6–9 These barriers included shorter hospital
length of stay, uncertainty regarding drug history and
previous medication experiences and the fact that most
patients do not have clinically stable heart failure during
their inpatient presentation.8–10 For most CHF patients,
hospitalization can be the beginning of medication optimization, which requires continuation after discharge.7
Although CHF is a significant proportion of hospital
admissions, general practitioners (GPs) manage this condition less frequently.11 Perceived barriers to the implementation of best practice guidelines for the management
of CHF in primary care include inadequate provision of
documentation from medical specialists, especially for
recently hospitalized patients, as well as a lack of information and awareness about the dosing of ACE inhibitor
and beta-blockers.8,9
Unfortunately, even with a strong evidence base, only
8–11% of all CHF patients access ‘specialized’ multidisciplinary CHF management.6 However, a hospital admission
within this group of patients does provide an opportunity
for specialist review and follow-on communication to the
GP on a recommended treatment plan and future goals. To
date, there has been a paucity of published work to show
how often this occurs in practice. The aim of this study was
to determine the level of documentation of management
plans and pharmacotherapy treatment goals, in CHF
patients discharged from a tertiary referral hospital to the
care of their GP, who would not be followed up by a
specialist heart failure programme.
Methods
A retrospective practice audit was conducted within the
general medical units of the Royal Adelaide Hospital
(RAH), a 650-bed tertiary referral institution located in
the city of Adelaide, South Australia.
Inclusion criteria
All patients discharged home to metropolitan Adelaide
(post codes 5000–5199) with International Classification
of Diseases-10 codes relating to a principal diagnosis of
CHF including ‘congestive heart failure’, ‘cardiomyopathy unspecified’, ‘ischaemic cardiomyopathy’, ‘left
ventricular failure’, ‘dilated cardiomyopathy’ and ‘acute
pulmonary oedema’ in the 12-month period commencing 1 January 2005 to 1 January 2006. These codes have
previously been shown to have a high specificity for the
Framingham criteria diagnosis of CHF.5 During this
period there was no specialized heart failure programme
available at the RAH.
596
Exclusion criteria
Patients discharged to nursing homes, other hospitals or
hospices, those deemed to be palliative for reasons other
than cardiac failure and patients with severe dementia
were excluded.
As part of the standard patient care process, the hospital
posts a discharge summary, including a medication list to
the patient’s GP, which is available electronically. Each of
these electronic summaries was reviewed by the research
team to assess: the dosage of ACE inhibitor, beta-blocker
and diuretic medication compared with the recommended
dosages, the presence of recommendations for ongoing
pharmacological and non-pharmacological management, arrangements for specialist follow up, total
number of comorbidities and regular medications prescribed on discharge.12–14 For each medication, documentation of contraindications was also sought within
the electronic discharge summary. For ACE inhibitors,
these included the presence of bilateral renal artery
stenosis, severe renal failure, hypotension, history of
angioedema or previous documented intolerance. For
beta-blockers these included hypotension, severe bradycardia or heart block, reversible airway disease or severe
chronic obstructive airway disease. Auditmaker (The
Australian Centre for Evidence-Based Clinical Practice,
Adelaide, South Australia) was used as the clinical audit
tool.15
In our assessment of the prescribed medications, we
took a generous approach towards the choice of drugs in
each class. Beta-blockers not specifically indicated for
CHF (atenolol and standard formulation of metoprolol)
were deemed to be acceptable, as many patients have
heart failure with preserved systolic function and these
agents may be used for rate control and improved diastolic filling.14 As the discharge summaries did not have
details of certain clinical parameters, which may have
limited drug dosage maximization, for example, blood
pressure, heart rate, divided the drug doses as being less
than half of the maximal dosage, which are usually starting doses, and more than or equal to half of maximal
dosage, suggesting that some dosage maximization had
already occurred.
Analysis
Data analysis and statistical measures were carried out
using MICROSOFT EXCEL 2003 and MICROSOFT ACCESS
2003 (Microsoft Corporation, Seattle, Washington, USA).
Descriptive statistics describing demographics and audit
outcomes are presented as proportions or means,
medians and standard derivations with 95% confidence
intervals.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Failure to communicate
Results
Table 1 Characteristics of patients reviewed
During the study period between 1 January 2005 and 1
January 2006, there were 351 admissions with the principal diagnosis of CHF. Of these there were 27 deaths
during admission, 31 patients were discharged to other
hospitals and 55 to residential care facilities. Of the
remaining 238 patients, 22 involved discharges outside of
the metropolitan area and 4 patients had either advanced
dementia or were palliative resulting in a total of 212
discharges, which were reviewed. The characteristics of
these patients are presented in Table 1. Patients were
elderly comorbid group, which was prescribed a median
of eight regular medications on discharge. Most patients
(60%) were followed up exclusively by their GPs with no
further specialist appointments. Discharge letters were
sent a median of 4 days after discharge with 25% of GPs
being sent the information 10 or more days after discharge. No discharge letter was sent in 9.4% (20) of
cases.
As can be seen from Table 2 only six of the 212 discharge letters (2.8%) had documented recommendations regarding future titration of ACE inhibitors. If the
patients who may not have benefited from up-titration
are excluded from this group (i.e. those with documented contraindication or those already prescribed
greater than half the recommended maximum dosage)
then approximately 5.6% (eight) patients had a documented dosage recommendation in their discharge
letter. Similarly, only 6.6% (14) of all letters had information regarding the future aims of the beta-blocker
therapy (Table 2). Again, if patients who may not have
benefited from up-titration of their beta-blockers are
excluded from this group, notation of future management was recorded in only 10% of discharge letters
(Table 2). Only six letters (2.8%) had information on
future dose alterations of diuretics. Recommendations
for non-pharmacological management, for example,
diuretic action plans, regular weight monitoring and
exercise plans were not found in any letter reviewed in
this audit.16
Characteristics
Discussion
This review of discharge communication is, to the best of
our knowledge only the second project to examine this
issue specifically in relation to CHF.
A similar study, carried out by Raval et al. also
stressed significant deficiencies in discharge documentation.17 Van Walraven et al. found that the provision of a
discharge letter decreased the rate of readmission
among patients admitted to hospital with acute medical illnesses.18 Inadequate documentation has been
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Age in years (mean and SD)
Sex n (%)
Male
Female
Length of stay – median (days)
Mean no. comorbidities
No. days from discharge to
posting of information to GP
Discharge information sent >10 days
after discharge (%)
Mean, n = 212
78.4 years (SD ⫾ 8.6)
105 (49.5)
107 (50.5)
6 days
5.43 (SD ⫾ 2.4)
7.75 (SD ⫾ 12.03),
median 4 days
53 (25.0)
GP, general practitioner; SD, standard deviation.
Table 2 Documentation of ACE inhibitor/angiotensin II antagonist and
beta-blocker prescription and dosage in discharge summary/letters
Documentation of ACE inhibitor/
angiotensin II antagonist prescription and
dosage within discharge summary letters
n (%) (95%CI),
n = 212
Total prescribed ACEI 156 (73.5%)
55 (25.9) (20–32)
High dose (dosage at ⱖ50% of
maximum dose†)
Low dose with recommendations for
8 (3.8) (2–7)
up-titration (<50% of maximum
dosage)
Low dose without recommendation
93 (43.9) (37–51)
for up-titration (<50% of
maximum dosage)
13 (6.1) (4–10)
Drug not prescribed with
documented contraindication†
Drug not prescribed without
23 (10.8) (7–16)
explanation
Discharge summary letter never written
20 (9.4) (6–14)
Documentation of beta-blocker prescription and dosage in discharge
letters
Total prescribed beta-blocker 93 (43.6%)
33 (16.0) (11–21)
High dose (dosage at ⱖ50% of
maximum dose†)
Low dose with recommendations
7 (3.3) (2–7)
for up-titration (<50% dosage)
Low dose without recommendations
53 (25.0) (19–31)
for up-titration (<50% of
maximum dosage)
32 (15.1) (11–20)
Drug not prescribed with
documented contraindication†
Not prescribed with recommendation
7 (3.3) (2–7)
for future prescription
Drug not prescribed without
54 (25.5) (20–31)
explanation
Letter not written
20 (9.4) (6–14)
Other, for example, prescribed sotalol
5 (2.4) (1–5)
for arrhythmia
†
No dosage recommendation required. 95%CI, ninety-five per cent confidence interval; ACEI, angiotensin-converting enzyme inhibitors.
597
Shakib et al.
identified as a barrier to the implementation of best
practice guidelines for CHF5 and as ideal management
of CHF reduces mortality and morbidity,3 it is possible
that improving discharge communication could improve
outcomes in patients with CHF.15 Overall, our review
showed that the quality of the documentation of
recommendations for pharmacological and nonpharmacological management was very poor. In particular, very few letters (<15%) outlined a plan of
management for the future dose titration of any of the
heart failure medications. The prescription of ACE
inhibitors and beta-blockers, both of which have proven
benefit in reducing heart failure mortality and morbidity, was not in keeping with the current best practice
guidelines13 and 70% of patients were receiving an
ACE inhibitor and 50% were receiving a beta-blocker.
This rate of prescription was, however, similar to those
reported from the CASE study (19) (ACEI 58% and
beta-blocker 14%) and The EuroHeart Survey (ACEI
61.8% and beta-blocker 36.9%) (14).
Given the fact that patients were admitted with decompensated heart failure (in the context of numerous
other comorbidities), this observation is perhaps not surprising. However, given that most were followed up by
their GPs and advice regarding this optimization was not
communicated in the discharge letter, it is unlikely that
this implementation gap will be bridged. Furthermore,
the discharge letter is usually the only routine tool for
communication with GPs. As CHF is a common condition within medical services of a large hospitals, but
constitutes only a small percentage (2.9%) of patients
seen by GPs,1,19 it may be unreasonable to expect GPs to
initiate optimal CHF management without specialist
guidance.9,13,19 The discharge letter may be a small
window of opportunity for education on the future treatment and drug optimization requirements in this group
of complex patients.20
The reasons for our findings are probably multifactorial
and their detailed investigation is beyond the scope of
this paper. At our institution, the discharge summary is
completed by the intern, who is the most junior member
of the treating team, and the one who is least likely to
appreciate the ongoing management requirements of the
patient. Our institution also does not have a chronic
disease management service and hence the chronic management of the patients is frequently not considered
during their inpatient management.
In light of our results we would recommend that
patients with chronic diseases, such as CHF, who have
acute hospital admissions and specialist contact punctuating their disease journey, have standardized recommendations for future management as an obligatory
component of their discharge planning process.21
598
Limitations
This review was conducted in only one institution and
reviewed information only from discharge letters. The
possible reasons why the discharge letters failed to
convey adequate documentation are numerous and we
could not assess other communications such as telephone
contact. The maximal doses of the medications for each
patient should ideally have been assessed by reviewing
the patient’s case notes to assess factors that would have
limited further dose increases. As this information was
not available in the discharge letters, we arbitrarily chose
a cut-off of 50% of the maximum recommended dosage
for each drug, to indicate that an adequate dose had been
prescribed. It is possible that this approach overestimated
the number of patients who would have benefited from
recommendations regarding dosage increases in their discharge letter. However, given the very poor level of documentation of dosage recommendation, this is unlikely to
have changed the overall conclusion of the study.
The strengths of the study were that the audit could
evaluate a large number of discharge letters for an entire
year across five different medical units of a large metropolitan hospital, hence sampling bias was eliminated.
Conclusion
The findings of this study suggest that even when there
are standardized approaches for discharge communication, for example, software for electronic discharge
letters, important management and treatment goals are
not included. This would suggest that other structured
approaches are required to implement a more satisfactory
transition for patients from the hospital to the community, especially for patients with chronic diseases.
Hospital discharge is an opportunity to communicate
effectively a management plan for treatment optimization to the patient’s principal clinician, and although
this opportunity is spurned, implementation gaps in the
management of CHF will probably remain.
References
1 Abhayaratna WP, Smith WT, Becker NG, Marwick TH,
Jeffery IM, McGill DA. Prevalence of heart failure and
systolic ventricular dysfunction in older Australians: The
Canberra Heart Study. Med J Aust 2006; 184: 151–4.
2 Stewart S. Prognosis of patients with heart failure
compared with common types of cancer. Heart Fail Monit
2003; 3: 87–94.
3 Levy D, Kenchaiah S, Larson MG, Benjamin EJ,
Kupka MJ, Ho KK et al. Long – term trends in the
incidence and survival with heart failure. N Engl J Med
2002; 347: 1397–1402.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Failure to communicate
4 Cline CMJ, Israelsson BYA, Willenheimer RB, Broms K,
Erhardt LR. Cost effective management programme for
heart failure reduces rehospitalisation. Heart 1998; 80:
442–6.
5 Dundon B, Shakib S, Thomas J, Maddison J, Philpot A.
Clinical trials to clinical practice in congestive cardiac
failure. Conference Proceedings from ASCEPT. 2002 Nov;
Melbourne; 135.
6 Clark RA, Eckert K, Stewart S, Phillips SM, Yallop JJ,
Tonkin AM et al. Rural and urban differentials in primary
care heart failure management: new data from the CASE
study. Med J Aust 2007; 186: 441–5.
7 Cleland JGF, Swedberg K, Follarh F, Komajda M,
Cohen-Solal A, Aguilar JC et al. The EuroHeart Failure
Survey Programme: survey on the quality of care among
patients with heart failure in Europe. Part 1: patient
characteristics and diagnosis. Eur J Heart Fail 2003; 24:
422–63.
8 Fuat A, Hungin APS, Murphy JJ. Barriers to accurate
diagnosis and effective management of heart failure in
primary care: quantitative study. Br Med J 2003; 326: 196.
9 Phillips SM, Marton RL, Tofler GH. Barriers to diagnosing
and managing heart failure in primary care. Med J Aust
2004; 181: 78–81.
10 Cleland JGF. Improving patient outcomes in heart
failure: evidence and barriers. Heart 2000; 84 Suppl I:
i8–10.
11 Phillips SM, Davies JM, Toffler GH. NICS heart failure
forum; improving outcomes in primary care. Med J Aust
2004; 181: 297–9.
12 Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ,
Ponikowski P, Poole-Wilson PA et al. ESC Guidelines for
the diagnosis and treatment of acute and chronic heart
failure 2008. The Task Force for the Diagnosis and
Treatment of Acute and Chronic Heart Failure 2008 of
the European Society of Cardiology. Developed in
collaboration with the Heart Failure Association of the
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
13
14
15
16
17
18
19
20
21
ESC (HFA) and endorsed by the European Society of
Intensive Care Medicine (ESICM). Eur J Heart Fail 2008;
10: 933–89.
Krum H, Jelinek M, Stewart S, Sindone A, Atherton JJ,
Hawkes AL et al. Guidelines for the prevention, detection
and management of people with chronic heart failure in
Australia. Med J Aust 2006; 185: 549–56.
Commonwealth Government of Australia. Therapeutic
Guidelines – Cardiovascular 2003. 2007 [cited 2006
Jan 1]. Available from URL: http://www.tg.com.au/
Shakib S, Phillips PA. The Australian Centre for
Evidence-Based Clinical Practice generic audit tool: Audit
maker for health professionals. J Eval Clin Pract [cited
2008 May] 2003; 9: 259–63. Available from URL: http://
www.auditmaker.org
Australian Medicines Handbook. Australian Medicines
Handbook. [Online edition] 2007 [cited Jan 26]. Available
from URL: http://www.amh.net.au
Raval AN, Marchiori GE, Arnold JM. Improving the
continuity of care following discharge of patients
hospitalized with heart failure: is the discharge summary
adequate? Can J Cardiol 2003; 19: 365–70.
Van Walraven C, Seth R, Austin P, Laupacis A. Effect of
discharge summary availability during post-discharge
visits on hospital readmission. J Gen Intern Med 2002; 17:
186–92.
Krum H, Tonkin AM, Currie R, Djundjek R, Johnston CI.
Chronic heart failure in Australian general practice. The
Cardiac Awareness Survey and Evaluation (CASE) Study.
Med J Aust 2001; 174: 439–44.
Rich MW. Heart Failure in the oldest patients: the impact
of comorbid conditions. Am J Geriatr Cardiol 2005; 14:
134–41.
Krumholz HM, Parent EM, Tu N, Vaccarino V, Wang Y,
Radford MJ et al. Readmission after hospitalisation for
congestive heart failure among medicare beneficiaries.
Arch Intern Med 1997; 157: 99–104.
599
Internal Medicine Journal 39 (2009) 600–605
O R I G I N A L A RT I C L E
Contributors to cognitive impairment in congestive heart failure:
a pilot case–control study
imj_1790
600..605
C. Beer,1,2 E. Ebenezer,3 S. Fenner,4 N. T. Lautenschlager,1,4,5 L. Arnolda,2 L. Flicker1,2
and O. P. Almeida1,4,5
1
Western Australia Centre for Health and Ageing, 2School of Medicine and Pharmacology and 5School of Psychiatry and Clinical Neurosciences,
University of Western Australia and 4Department of Psychiatry, Royal Perth Hospital, Perth, Western Australia, Australia, and 3Department of
Psychological Medicine, University of Malaya, Kuala Lumpur, Malaysia
Key words
cardiac failure, heart failure, memory,
cognition, cognitive function.
Correspondence
Christopher Beer, Western Australia Centre for
Health and Ageing (M573), University of
Western Australia, 35 Stirling Highway,
Crawley, WA 6009, Australia.
Email: [email protected]
Received 24 April 2008; accepted
17 June 2008.
doi:10.1111/j.1445-5994.2008.01790.x
Abstract
Background: Cognitive impairment and heart failure are both serious health
problems related to population ageing. Impaired cognitive function is an
important but underrecognized complication of congestive heart failure
(CHF). The aim of the study was to examine the sociodemographic, clinical,
neuroimaging and biochemical parameters affecting cognition in CHF.
Methods: Thirty-one patients with CHF (left-ventricular ejection fraction <
40%) and 24 controls without CHF, all free of clinically significant cognitive
impairment, participating in a case–control study were assessed using a cognitive battery (CAMCOG), a depression scale, 6-min-walk test, left-ventricular
ejection fraction, semi-quantitative magnetic resonance imaging, and cortisol,
aldosterone and renin concentrations.
Results: The CHF patients had lower CAMCOG scores than controls (93.5 ⫾
6.1 vs 99.9 ⫾ 2.4, P < 0.001) and had significantly lower scores on visuospatial,
executive function, visual memory and verbal learning tasks. Concentrations
of renin and aldosterone were higher in patients with CHF (5.4 ⫾ 6.0 vs 0.8
⫾ 0.7 mU/L, P < 0.001 and 598.2 ⫾ 306.2 vs 346.0 ⫾ 201.5, P = 0.003). Right
medial temporal lobe atrophy was more prominent in CHF (P = 0.030). Left
medial temporal lobe atrophy and deep white matter hyperintensities showed
moderate association with cognitive scores in CHF, whereas functional capacity and biochemical parameters were fairly correlated to cognition.
Conclusion: Congestive heart failure is associated with a pattern of generalized cognitive decline. Structural brain changes, functional capacity and biochemical parameters are associated with the cognitive performance of patients
with CHF, but their contribution appears modest. The design of a definitive
case–control study is described.
Introduction
Congestive heart failure (CHF) is a common complication
of most diseases of the heart. As the population ages, costs
due to CHF will rise substantially, as many of these diseases
Funding: The study was supported by a project grant from the
National Health and Medical Research Council of Australia
(project number 403996).
Conflict of interest: None.
600
are age dependent. A 75% increase in costs due to CHF is
projected in the USA between 2000 and 2050.1 CHF is
associated with increased mortality and significant morbidity with median survival among older adults of less
than 5 years.2
An important but underrecognized complication of
CHF is impaired cognitive function. Two recent systematic reviews concluded that CHF is associated with deficits
of multiple cognitive domains, including memory, concentration, attention shifting and psychomotor speed.3,4
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cognitive impairment in heart failure
There is also evidence that patients with appreciable cognitive deficits have greater long-term morbidity and mortality than those with normal cognitive function, which
suggests that impaired mental function may worsen the
natural course of CHF.5–7 The reverse, exacerbation of
cognitive impairment by CHF, may also be true. This
possibility is supported by the observation that CHF
increases the risk of dementia and Alzheimer’s disease in
later life.8,9
The factors that mediate the interaction between
impaired cardiac function and cognitive decline may be
varied, with possible candidates including cerebrovascular disease and its associated risk factors, decreased
cardiac output and cerebral blood flow, hypotension and
neurohormonal imbalance.10–12 Nonetheless, there is
limited systematic information on how these factors
interact to hinder brain function in persons with CHF.
We designed the present study to investigate the contribution of sociodemographic, clinical, neuroimaging
and biochemical parameters to the cognitive scores of
adults with CHF to inform design of a larger case–control
study.
Methods
Subjects
Patients with stable New York Heart class II CHF were
recruited from the Departments of Cardiology and
Internal Medicine of the Royal Perth Hospital. Only
those with ejection fractions of 40% or less on transthoracic echocardiography were included in this study
group. Healthy controls were recruited through advertising in the local media; they had no symptoms of CHF
and had an ejection fraction greater than 40%. All participants with prior history of stroke or current MiniMental State Examination (MMSE) score lower than 24
(indicative of clinically significant cognitive impairment)
were excluded.13
Assessment and procedures
We used section A of the Cambridge Examination for
Mental Disorders of the Elderly (CAMDEX) to collect
basic demographic and clinical information from participants.5 The cognitive assessment was designed to test a
broad range of cognitive domains and included section B
of the CAMDEX (CAMCOG), which provides a summary
measure of general intellectual ability, the California
Verbal Learning Test14 and Brief Visuospatial Memory
Test – Revised,15 which assess verbal and visual memory,
tests of verbal fluency (using the letters ‘F’, ‘A’ and ‘S’ and
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Block Design,16 which offers a measure of executive function and visuospatial ability, respectively.
After completing the cognitive examination, participants were asked to rate the Living with Heart Failure
Questionnaire17 and the Even Briefer Assessment Scale
for Depression (EBAS-DEP).18 Finally, subjects underwent the 6-min-walk test.19
Fasting blood samples were drawn between 07.00 and
08.30 hours on the day of the assessment for evaluation
of cortisol (nmol/L), aldosterone (pmol/L) and renin
(mU/L) concentrations. The coefficient of variation for
these assays is approximately 5%.
Imaging
Eligible patients who consented underwent magnetic
resonance imaging (MRI), using a 1.5-T Siemens Scanner
(Erlangen, Germany). For each subject, 176 0.9-mm sagittal slices with in-plane resolution of 1 mm ¥ 1 mm were
acquired. T1-weighted images were rated by a trained
member of the research team (S. F., an experienced clinician) who was blind to the identity and clinical status of
subjects. Sulcal widening (SW) and ventricular enlargement (VE) were rated by comparison to a standard reference image that showed moderate brain changes. A rating
of ‘0’ indicated no SW or VE, ‘1’ indicated changes less
marked than in the reference image, ‘2’ indicated changes
of a similar degree to the reference image and ‘3’ indicated
changes in excess of those in the reference image. Previous
studies have shown that this semi-quantitative rating of
brain images is both reliable and valid.20–22 Medial temporal atrophy left and right (MTA:L and MTA:R) were rated
using the method described by Scheltens et al.,23–25 in
which hippocampal slices are examined for widening of
the choroid fissure, enlargement of the temporal horn of
the lateral ventricle and decreased height of the hippocampus (0 absent, 1 minimal, 2 mild, 3 moderate and 4
severe atrophy). Periventricular white matter changes
(PVWMC) and deep white matter changes (DWMC) were
rated using the method originally described by Fazekas
et al.26 For PVWMC, a rating of ‘0’ indicated no changes,
‘1’ indicated ‘caps’ or pencil-thin periventricular bands,
‘2’ indicates a smooth ‘halo’ and 3 indicated irregular
periventricular changes extending into deep white matter.
For DWMC ‘0’ indicated no changes, ‘1’ indicated punctate foci, ‘2’ beginning confluence of foci and ‘3’ large
confluent areas.
Before examination of study scans, intra-rater reliability was established by repeated rating of a set of 15 scans.
Each scan was rated twice in random sequence. Kappa
scores were 0.9 for VE, 0.7 for SW, 0.7 for MTA:L, 0.6 for
MTA:R, 0.6 for PVWMC and 0.8 for DWMC. This degree
of reliability was considered acceptable and within the
601
Beer et al.
Results
range of previously published figures for reliability of
visual rating methods.27
Clinical, laboratory and imaging results in
patients with CHF compared to controls
Statistical analysis
Thirty-one patients and 24 controls entered the study.
Demographic, clinical features cognitive outcomes and
laboratory values for both groups are summarized in
Table 1. Subjects with CHF had a mean ejection fraction
of 25.7 ⫾ 8.1%. Subjects with CHF had higher scores on
the Living with Heart Failure Questionnaire and walked
a shorter distance on the 6-min-walk test.
Patients with CHF had a mean CAMCOG score 6.4
points lower than that of the control group (P < 0.001).
Table 1 shows that patients with CHF had significantly
lower scores than controls across most of the tasks used,
including visuospatial, executive function, visual memory
and verbal learning tasks. Scores on the Even Briefer
Assessment Scale for Depression were not significantly
different between patients with CHF and control subjects.
Concentrations of renin and aldosterone were higher
in patients with CHF compared with control subjects and
Data were managed and analysed with SPSS for windows,
version 12 (SPSS, Chicago, IL, USA). We used graphic
methods to examine the distribution of data. We used the
Pearson c2-statistic to analyse the distribution of categorical variables and Student’s t-test for between-group comparisons of numerical, normally distributed data. Mann–
Whitney rank test was used for the between-group
comparison of ordinal data. We used Pearson correlation
coefficients to determine the strength of the association
between cognitive scores and variables of interest in
patients with CHF and partial correlations to take the
effect of age, sex and education into account.
The study was approved by the Royal Perth Hospital
Human Research Ethics Committee and all participants
provided written informed consent. The study conforms
to the principles outlined in the Declaration of Helsinki.
Table 1 Characteristics and clinical features of patients with CHF and control
Sociodemographic and clinical measures
Age, mean years (SD)
Male sex, n (%)
Age left school, mean years (SD)
6-min-walk test, mean metres (SD)
EBAS-DEP, mean score (SD)
Cognitive measures
CAMCOG, mean score (SD)
CVLT total recall, mean score (SD)
CVLT short delay, mean score (SD)
CVLT long delay, mean score (SD)
CVLT recognition, mean score (SD)
BVMT total recall, mean score (SD)
BVMT delayed recall, mean score (SD)
BVMT recognition, mean score (SD)
Verbal fluency (letters ‘F,’, ‘A’ and ‘S’) mean number of words (SD)
Block Design, mean score (SD)
Biochemical measures
Cortisol concentration, mean mmol/L (SD)†
Aldosterone concentration, mean pmol/L (SD)†
Renin concentration, mean mU/L (SD)†
MRI measures
Ventricular enlargement, mean score (SD)‡
Sulcal widening, mean score (SD)‡
Left medial temporal lobe atrophy, mean score (SD)
Right medial temporal lobe atrophy, mean score (SD)
Periventricular white matter changes, mean score (SD)
Deep white matter changes, mean score (SD)
CHF (n = 31)
Controls (n = 24)
Statistic
P
54.3 (10.6)
26 (83.9)
16.1 (1.9)
492.8 (91.4)
1.2 (1.6)
56.1 (8.2)
20 (83.3)
16.6 (2.2)
584.6 (57.3)
1 (1.8)
t = -0.70
c = 0.00
t = -0.89
t = -4.29
z = 1.22
0.488
0.957
0.378
<0.001
0.222
93.5 (6.1)
44.1 (10.1)
8.9 (3.1)
9.8 (3.1)
14.5 (1.5)
20.6 (6.3)
8.2 (2.8)
5.8 (0.4)
35.6 (9.0)
38.2 (11.2)
99.9 (2.4)
52.5 (8.2)
11.0 (2.7)
11.8 (2.5)
15.2 (1.0)
24.2 (6.0)
9.8 (2.2)
5.9 (0.3)
43.9 (8.6)
47.2 (9.4)
t = -4.82
t = -3.32
t = -2.61
z = -2.63
z = -1.93
t = -2.16
z = -2.24
z = -1.40
t = -3.45
t = -3.17
<0.001
0.002
0.009
0.008
0.054
0.035
0.025
0.160
0.001
0.002
510.7 (151.6)
598.2 (306.2)
5.4 (6.0)
441.8 (158.7)
346.0 (201.5)
0.8 (0.7)
t = 1.50
t = 3.10
z = 3.92
0.141
0.003
<0.001
1.2 (0.8)
1.5 (0.5)
0.3 (0.5)
0.2 (0.4)
0.8 (0.6)
1.5 (0.6)
1.4 (0.8)
1.5 (0.6)
0.1 (0.4)
0 (0)
1.1 (0.4)
1.7 (0.7)
z = -1.06
z = 0.00
z = 0.86
z = 2.14
z = -2.11
z = -1.13
0.287
1.000
0.388
0.030
0.034
0.259
†
Laboratory data were available for 30 participants with CHF and 18 controls. ‡MRI scans were available for 19 participants with CHF and 20 controls.
t, t-statistic from Student’s t-test; z, standardized statistic value for Mann–Whitney test. BVMT, Brief Visuospatial Memory Test – Revised; CHD, congestive
heart failure; CVLT, California Verbal Learning Test; EBAS-DEP, Even Briefer Assessment Scale for Depression; MRI, magnetic resonance imaging; SD,
standard deviation.
602
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cognitive impairment in heart failure
there was a trend towards higher levels of cortisol in
subjects with CHF compared with controls.
Brain images were available and rated for 19 of 31 with
CHF and 20 of 24 controls. The low number of magnetic
MRI scans available related to contraindications to MRI
scanning (most commonly claustrophobia, use of pacemaker or presence of other implanted metal objects).
Visual rating (Table 1) suggested that right medial temporal lobe atrophy was statistically significantly more
prominent in patients with CHF compared with controls.
However, periventricular white matter hyperintensities
were found to be less severe in patients with CHF than
controls. Visual rating failed to show significant differences in the other parameters examined between
patients with CHF and control subjects (Table 1).
Correlates of cognitive impairment
among CHF patients
Correlations between CAMCOG scores (primary cognitive measure of our study) and clinical, biochemical and
imaging variables among patients with heart failure are
summarized in Table 2. Moderate negative correlations
were found with left medial temporal lobe atrophy and
severity of deep white matter hyperintensities. Fair correlations were found between total CAMCOG score and
6-min-walk test distance. Higher concentrations of cortisol correlated fairly to worse cognitive performance,
whereas activation of the renin–aldosterone system
showed fair positive correlation to cognitive performance. Sulcal widening, right medial temporal lobe
atrophy and periventricular white matter hyperintensities, all had fair negative correlations to total CAMCOG
Table 2 Correlation coefficients between CAMCOG score and clinical,
biochemical and imaging variables among participants with heart failure
6-min-walk test
Left-ventricular ejection fraction
EBAS-DEP score
Cortisol concentration
Aldosterone concentration
Renin concentration
Cerebral ventricular enlargement
Sulcal widening
Left medial temporal lobe atrophy
Right medial temporal lobe atrophy
Periventricular hyperintensities
Deep white matter hyperintensities
Correlation
coefficient†
Adjusted
coefficient‡
0.42*
0.05
-0.24
-0.21
0.21
0.19
0.05
-0.21
-0.50*
-0.28
-0.15
-0.43
0.32
0.15
-0.13
-0.30
0.23
0.29
0.06
-0.31
-0.47
-0.31
-0.29
-0.46
*P < 0.05. †Pearson correlation coefficient. ‡Adjusted for age, gender and
education. Strength of the association: <0.20 poor, 0.21–0.40 fair, 0.41–
0.60 moderate, 0.61–0.80 good, >0.80 very good. EBAS-DEP, Even Briefer
Assessment Scale for Depression.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
score. Ejection fraction, score on the EBAS and VE were
only poorly correlated to cognitive performance.
Discussion
The results of this study confirm that even after exclusion
of participants with MMSE <24, CHF is associated with a
pattern of generalized cognitive impairment relative to
controls. It is very likely that this difference of 6 points on
the CAMCOG scores (and 1.25 points on MMSE) is clinically meaningful (being more than 2 standard deviations
lower than controls). These differences probably translate
into clinically important differences in everyday activities
including coping with the requirements of a serious
chronic illness.
In addition, subjects with CHF showed structural brain
changes that were associated with general cognitive
dysfunction, although these failed to consistently discriminate patients from controls. Our inability to
show significant differences on structural brain imaging
between CHF patients and controls may be because of
either the relative insensitivity of these techniques or the
low power of the study with respect to these end-points.
Alternatively, the selection of participants who were free
of significant cognitive deficits may have biased the
results towards the null hypothesis because of the selection of patients with a relatively well-preserved brain.
Although possible differences in the degree of right-sided
and left-sided structural brain changes in patients with
CHF are interesting, the present data may not be reliable.
Given the lack of a biologically compelling hypothesis to
explain differences between left-sided and right-sided
changes, even in larger datasets, authors have tended to
suggest that observed differences may be spurious and
attributable to methodological factors.
We also detected markers of neurohormonal activation
in patients with CHF, who had higher levels of cortisol,
aldosterone and renin than controls. We found a fair
inverse association between cognitive scores and the
serum concentration of cortisol and a fair direct correlation between cognitive scores and levels of aldosterone
and renin. The direct association between aldosterone
levels and cognitive performance was unexpected. Previous data suggested that higher levels of both cortisol
and aldosterone are associated with increased mortality
among patients with CHF.28 Because of our small sample
size this unexpected result should be interpreted with
caution.29 In addition these data were exploratory, rather
than to test an a priori hypothesis. Drug therapies for
CHF also complicate interpretation of these data because
many commonly used heart failure therapies (including
beta-blockers, angiotensin-converting enzyme inhibitors,
603
Beer et al.
angiotensin receptor antagonists and spironolactone)
affect the renin–angiotensin system levels.
The results also show that functional measures of CHF,
such as left-ventricular ejection fraction and the 6-min walking test have only a modest direct association with cognitive
function. Ejection fraction was previously found to be independently associated with cognitive function, but this may
have been because of the inclusion of healthy controls in
those analyses.30 Thus, the poor correlation between ejection fraction and cognitive function in the present data is of
interest. These data suggest that ejection fraction may be
a less important contributor to cognitive dysfunction in
subjects with stable CHF than previously thought. Further
data are required to clarify this association.31
The study results cannot be adequately explained by
confounding because of age, sex, schooling or presence of
depressive symptoms. However, interpretation of our
results is limited by the fact that we did not collect data
on other factors that may have led to cognitive decline in
this patient group, such as smoking,32 diabetes,33 hypertension,34 hypotension,12,35 cerebral hypoperfusion,36
decreased physical and mental activity and other potentially relevant factors, such as pulse and the presence
and degree of orthostatic hypotension.37 Although cerebrovascular disease and significant cerebral hypoperfusion are relatively unlikely to confound the observed
relationship, given the age of our subjects and exclusion
of subjects with severe heart failure, further data will be
required to determine whether CHF is independently
associated with cognitive dysfunction after allowing for
the presence and extent of these risk factors.
This small study included relatively young, highly
functioning individuals with heart failure due to systolic
dysfunction. Further data are required to determine the
magnitude and range of cognitive dysfunction associated
with CHF in older, frailer populations in routine clinical
practice and in patients with diastolic dysfunction. Determining the factors that predict the susceptibility of an
individual patient with CHF to develop cognitive dysfunction will also be of interest. Longer-term observational data are also required to determine whether the
cognitive abnormalities associated with CHF are stable
over time and can be improved by treatment that
improves cardiac function or that minimizes the negative
effect of other relevant risk factors or comorbidities.
Despite these limitations, this study has shown substantial cognitive changes in visuospatial, executive, memory
and learning functions in patients with heart failure. The
data also suggest that numerous factors possibly contribute to the cognitive deficits observed among patients with
CHF. Further data are required to improve our understanding of how these factors interact to produce cognitive dysfunction over time in patients with CHF.
604
These data and conclusions have led to design of a
larger study, which is now under way with National
Health and Medical Research Council funding support. In
light of the difficulty adequately controlling for potential
confounding variables, the ongoing study includes two
control groups. The first control group comprises healthy
persons to control for the effects of normal ageing.
The second control group includes participants with
ischaemic heart disease (but preserved systolic function).
The latter are anticipated to have a risk-factor profile
similar to participants with CHF and thus adequately
control for the effects of vascular risk factors and comorbidities. Participants will be followed for 2 years to determine the association of heart failure with cognitive
decline over time. This study also includes an imaging
component and will provide more definitive data regarding the association of CHF with structural brain changes.
Acknowledgements
We would also like to thank the volunteers who participated in the study and our clinical colleagues who
assisted in recruiting volunteers.
References
1 Garrett N, Martini EM. The boomers are coming: a total
cost of care model of the impact of population aging on
the cost of chronic conditions in the United States. Dis
Manag 2007; 10: 51–60.
2 Thomas S, Rich MW. Epidemiology, pathophysiology,
and prognosis of heart failure in the elderly. Clin Geriatr
Med 2007; 23: 1–10.
3 Almeida OP, Flicker L. The mind of a failing heart: a
systematic review of the association between congestive
heart failure and cognitive functioning. Intern Med J
2001; 31: 290–95.
4 Vogels RL, Scheltens P, Schroeder-Tanka JM, Weinstein HC.
Cognitive impairment in heart failure: a systematic review
of the literature. Eur J Heart Fail 2007; 9: 440–49.
5 McLennan SN, Pearson SA, Cameron J, Stewart S.
Prognostic importance of cognitive impairment in chronic
heart failure patients: does specialist management make a
difference? Eur J Heart Fail 2006; 8: 494–501.
6 Stump TE, Callahan CM, Hendrie HC. Cognitive
impairment and mortality in older primary care patients.
J Am Geriatr Soc 2001; 49: 934–40.
7 Zuccala G, Onder G, Pedone C, Cocchi A, Carosella L,
Cattel C et al. Cognitive dysfunction as a major
determinant of disability in patients with heart failure:
results from a multicentre survey. On behalf of the GIFA
(SIGG-ONLUS) Investigators. J Neurol Neurosurg Psychiatry
2001; 70: 109–12.
8 Huynh BC, Rovner A, Rich MW. Long-term survival in
elderly patients hospitalized for heart failure: 14-year
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Cognitive impairment in heart failure
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
follow-up from a prospective randomized trial. Arch
Intern Med 2006; 166: 1892–8.
Qiu C, Winblad B, Marengoni A, Klarin I, Fastbom J,
Fratiglioni L. Heart failure and risk of dementia and
Alzheimer disease: a population-based cohort study. Arch
Intern Med 2006; 166: 1003–8.
Almeida OP, Tamai S. Clinical treatment reverses
attentional deficits in congestive heart failure. BMC
Geriatr 2001; 1: 2.
Gruhn N, Larsen FS, Boesgaard S, Knudsen GM,
Mortensen SA, Thomsen G et al. Cerebral blood flow in
patients with chronic heart failure before and after heart
transplantation. Stroke 2001; 32: 2530–33.
Zuccala G, Onder G, Pedone C, Carosella L, Pahor M,
Bernabei R et al. Hypotension and cognitive impairment:
selective association in patients with heart failure.
Neurology 2001; 57: 1986–92.
Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”.
A practical method for grading the cognitive state of patients
for the clinician. J Psychiatr Res 1975; 12: 189–98.
Delis D, Kramer J, Kaplan E, Ober B. California Verbal
Learning Test – Research Edition. San Antonio (TX): The
Psychological Corporation; 1987.
Benedict R. Brief Visuospatial Memory Test-Revised. Odessa
(FL): Psychological Assessment Resources, Inc; 1997.
Wechsler D. The Wechsler Adult Intelligence Scale III. San
Antonio (CA): Psychological Corporation, Harcourt
Brace; 1997.
Rector TS, Kubo SH, Cohn JN. Validity of the Minnesota
living with heart failure questionnaire as a measure of
therapeutic response to enalapril or placebo. Am J Cardiol
1993; 71: 1106–7.
Allen N, Ames D, Ashby D, Bennetts K, Tuckwell V,
West C. A brief sensitive screening instrument for
depression in late life. Age Ageing 1994; 23: 213–19.
Lipkin DP, Scriven AJ, Crake T, Poole-Wilson PA. Six
minute walking test for assessing exercise capacity in
chronic heart failure. Br Med J (Clin Res Ed) 1986; 292:
653–5.
Ghanem MH. CT scan in psychiatry. A review of the
literature. Encephale 1986; 12: 3–12.
Victoroff J, Mack WJ, Grafton ST, Schreiber SS, Chui HC.
A method to improve interrater reliability of visual
inspection of brain MRI scans in dementia. Neurology
1994; 44: 2267–76.
Yue NC, Arnold AM, Longstreth WT Jr, Elster AD,
Jungreis CA, O’Leary DH et al. Sulcal, ventricular, and
white matter changes at MR imaging in the aging brain:
data from the cardiovascular health study. Radiology
1997; 202: 33–9.
Scheltens P, Launer LJ, Barkhof F, Weinstein HC,
Jonker C. The diagnostic value of magnetic resonance
imaging and technetium 99m-HMPAO
single-photon-emission computed tomography for the
diagnosis of Alzheimer disease in a community-dwelling
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
24
25
26
27
28
29
30
31
32
33
34
35
36
37
elderly population. Alzheimer Dis Assoc Disord 1997; 11:
63–70.
Frisoni GB, Scheltens P, Galluzzi S, Nobili FM, Fox NC,
Robert PH et al. Neuroimaging tools to rate regional
atrophy, subcortical cerebrovascular disease, and regional
cerebral blood flow and metabolism: consensus paper of
the EADC. J Neurol Neurosurg Psychiatry 2003; 74:
1371–81.
Wahlund LO, Julin P, Johansson SE, Scheltens P. Visual
rating and volumetry of the medial temporal lobe on
magnetic resonance imaging in dementia: a comparative
study. J Neurol Neurosurg Psychiatry 2000; 69: 630–35.
Fazekas F, Chawluk JB, Alavi A, Hurtig HI,
Zimmerman RA. MR signal abnormalities at 1.5 T in
Alzheimer’s dementia and normal aging. AJR Am J
Roentgenol 1987; 149: 351–6.
Fan YH, Lam WW, Mok VC, Huang RX, Wong KS.
Variability and validity of a simple visual rating scale in
grading white matter changes on magnetic resonance
imaging. J Neuroimaging 2003; 13: 255–8.
Guder G, Bauersachs J, Frantz S, Weismann D, Allolio B,
Ertl G et al. Complementary and incremental mortality
risk prediction by cortisol and aldosterone in chronic
heart failure. Circulation 2007; 115: 1754–61.
von Bohlen und Halbach O, Albrecht D. The CNS reninangiotensin system. Cell Tissue Res 2006; 326: 599–616.
Zuccala G, Cattel C, Manes-Gravina E, Di Niro MG,
Cocchi A, Bernabei R. Left ventricular dysfunction: a
clue to cognitive impairment in older patients with heart
failure. J Neurol Neurosurg Psychiatry 1997; 63: 509–12.
Wahlund L-O, Julin P, Lindqvist J, Scheltens P. Visual
assessment of medial temporal lobe atrophy in demented
and healthy control subjects: correlation with volumetry.
Psychiatry Res Neuroimaging 1999; 90: 193–9.
Almeida OP, Hulse GK, Lawrence D, Flicker L. Smoking
as a risk factor for Alzheimer’s disease: contrasting
evidence from a systematic review of case-control and
cohort studies. Addiction 2002; 97: 15–28.
Xiong GL, Plassman BL, Helms MJ, Steffens DC. Vascular
risk factors and cognitive decline among elderly male
twins. Neurology 2006; 67: 1586–91.
Peila R, White LR, Masaki K, Petrovitch H, Launer LJ.
Reducing the risk of dementia: efficacy of long-term
treatment of hypertension. Stroke 2006; 37: 1165–70.
Nilsson SE, Read S, Berg S, Johansson B, Melander A,
Lindblad U. Low systolic blood pressure is associated
with impaired cognitive function in the oldest old:
longitudinal observations in a population-based sample
80 years and older. Aging Clin Exp Res 2007; 19: 41–7.
Pullicino PM, Hart J. Cognitive impairment in congestive
heart failure?: Embolism vs hypoperfusion. Neurology
2001; 57: 1945–6.
Lautenschlager NT, Almeida OP. Physical activity and
cognition in old age. Curr Opin Psychiatry 2006; 19:
190–93.
605
Internal Medicine Journal 39 (2009) 606–612
O R I G I N A L A RT I C L E
Selection of medical patients for prophylaxis of venous
thromboembolism based on analysis of the
benefit–hazard ratio
imj_1796
606..612
J. A. Millar
Department of Internal Medicine, Royal Perth Hospital and Royal Perth Hospital Unit, University of Western Australia School of Medicine, Perth,
Western Australia, Australia
Key words
benefit–harm ratio, low molecular weight
heparin, major bleeding, thromboprophylaxis,
treatment guideline, venous
thromboembolism.
Correspondence
J. Alasdair Millar, Medicine Department,
Southland District Hospital, PO Box 828,
Invercargill 9840, New Zealand.
Email: [email protected]
Received 8 April 2008; accepted 3 July 2008.
doi:10.1111/j.1445-5994.2008.01796.x
Abstract
Background: Medical patients may benefit from anticoagulant prophylaxis
of venous thromboembolism (VTE), but assessment of thrombotic risk is
complex. I describe a method for estimating the minimum thrombotic risk
required to ensure that a reasonable benefit–hazard ratio is maintained.
Methods: An equation was derived relating baseline VTE risk and a
minimum acceptable benefit–hazard ratio (R), defined as ‘pulmonary embolus
(PE) alone’, ‘PE or symptomatic proximal deep venous thrombosis (DVT)’, or
‘PE or any symptomatic DVT’ prevented per major bleeding. The equation was
used to estimate the relative risk (RR) of thromboembolism required for net
benefit (main outcome measure). The PREVENT study was the primary data
source, backed by data from two meta-analyses.
Results: For R ranging from 3 to 10, the RR required for net beneficial
prophylaxis was 6.5–21.6 (PE alone); 3.0–9.9 (PE or symptomatic proximal
DVT); and 2.3–7.6 (PE or any symptomatic DVT), respectively. These RR are
possible only in the presence of risk factors of high weighting. Sensitivity
analysis showed that the findings were robust to changes in baseline assumptions related to thrombosis and bleeding rates.
Conclusion: A method for risk assessment for medical thromboprophylaxis
has been developed. The results suggest that only a minority of medical
patients with high RR should receive prophylaxis.
Introduction
It has been proposed that the benefits of venous thromboembolism (VTE) prophylaxis seen after surgery should
be extended to medical patients and that a routine VTE
risk assessment should be undertaken with this end in
view.1 However, the precise basis for the proposed assessment is unclear. The level of risk requiring VTE prophylaxis, and hence the subset of patients who should be
treated, is undefined. Calculation of risk and determination of whether it exceeds a limit that satisfies the requirement for prophylaxis are complex processes because of
uncertainty over how clinical trial groups relate to a particular patient, variation of VTE risk according to the
primary medical condition, and additional comorbidities
that are additional risk factors of varying weight.
In this paper I describe an approach based on benefit–
hazard (B:H) ratio analysis that points to a simple and
practical means of diagnosing a VTE risk requiring prophylaxis in medical patients.
Methods
Assumptions
Funding: None.
Conflict of interest: None.
606
1 The approach is based on analysis of rates of pulmonary
embolus (PE) ⫾ symptomatic deep venous thrombosis
(DVT) and the extent to which these may be affected by
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Benefit–hazard ratio in medical VTE prophylaxis
Derivation of relationship between R and
relative risk of VTE
Consider the effect of LMWH on PE with and without
symptomatic DVT, and major bleeding, in a group of
medical patients. Figure 1 shows PE and DVT data from
the PREVENT study as an example.3 The effect of prophylaxis on each end-point is (A - B) and the increase in
the number with major bleeding due to the drug is (X Y). Under assumption 4(ii) (X - Y) is constant (k). Hence
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
1.2
A3
1
0.8
%
prophylaxis. According to the American College of Chest
Physicians guidelines, the main but not the sole priority of
VTE prophylaxis is prevention of PE.2 Hence, three endpoints were analysed in this study: ‘PE alone’, ‘PE or
symptomatic proximal DVT’ and ‘PE or any symptomatic
DVT’.
2 The reduction PE and symptomatic DVT risk with prophylaxis is the same as the reduction in the primary trial
end-points and approximates to a relative risk reduction
(RRR) of 0.48. This is the weighted average risk reduction
as reported in the PREVENT3 (0.45) and MEDENOX4
(0.63) studies. This assumption was warranted because
the low baseline PE rate and RRR in clinical trials that
were underpowered for clinical VTE end-points precludes
accurate measurement of the effect of prophylaxis on
PE and symptomatic DVT. This source of uncertainty
was investigated by repeating the analysis using data of
medical anticoagulant thromboprophylaxis from two
meta-analyses and by means of a sensitivity analysis.
Both meta-analyses included the PREVENT study, which
was used as the primary data source because (i) it is
the largest randomized study of prophylaxis with low
molecular weight heparin (LMWH) (n = 1850/group
versus 370 in MEDENOX), (ii) PREVENT study results
were not unduly affected by anticoagulant therapy
offered to patients in whom subclinical thrombosis was
detected by an investigative imaging technique before the
trial end-point day (in PREVENT, day 21 of admission)
and (iii) data on symptomatic VTE events were reported.
3 Only LMWH anticoagulant prophylaxis is considered.
Of the two possible candidate drug classes used in Australia (unfractionated heparin (UH) and LMWH), the evidence is in favour of the latter because of decreased risk
of bleeding and heparin-induced thrombocytopenia
(HITS) compared with unfractionated heparin.5,6 Accordingly, the applicable evidence base is randomized trials
containing usable data on the effect of LMWH prophylaxis on PE and symptomatic DVT.
4 (i) The hazard considered is major bleeding and (ii)
the risk of drug-induced major bleeding at a standard
prophylactic dose of LMWH is constant in the general
medical groups considered for prophylaxis.
A2
0.6
0.4
B3
X
B2
A1
B1
Y
0.2
0
No prophylaxis
Prophylaxis
Figure 1 Rates (%) of pulmonary embolism (PE alone; ), PE or proximal
symptomatic (PS) DVT (PE or PSDVT, ), PE or any symptomatic (AS) DVT
(PE or ASDVT; 䊐) and major bleeding ( ) as reported in the PREVENT
study2 of venous thromboembolism prophylaxis using dalteparin (prophylaxis) versus placebo (no prophylaxis). The labels A and B for each endpoint (numbered 1–3), X and Y have the same meanings as in the
derivation of Equation (2).
the marginal B:H ratio R = A - B/k (Eqn 1) thus kR = A B. However, B = A(1 - E) where E is the relative risk
reduction. Thus kR = AE and, as E is constant (0.48),
A=
k
R = k ′R
E
(2)
Equation 2 states that there is a direct relationship
between the baseline risk of a thrombotic end-point
deemed to be worthy of prophylaxis and the agreed
minimum ratio of benefit to harm, other things being
equal. Patients with a value of A less than the value given
by Equation 1 have a B:H ratio (R) that is less than the
agreed value and in consequence should not receive
thromboprophylaxis. I assumed that the minimum
acceptable value of R lies in the range 3–10.
The relative risk (RR) of the event in the subpopulation
defined by Equation 1, compared with the group defined
in relevant clinical trials, is obtained by division (A/Atrial)
where Atrial is the observed baseline risk of the event
(without prophylaxis) in clinical trials.
The final step is to compare the RR derived from the
analysis with the known risk factors for VTE in medical
patients and thereby determine which risk factors are
sufficient to discriminate for or against prophylaxis. For
example, if the relative risk given by A/Atrial = 6.0, then
this outcome would be satisfied, according to the risk
factors shown in Table 1, in patients who have a history
of DVT or VTE, in addition to the inclusion criteria of trials
such as PREVENT.7,8 The condition would also be satisfied
by the presence of two or more additional risk factors of
lesser weight, where the aggregate RR would be the
product of the individual relative risks. Patients whose
aggregate RR was less than 6.0 (in this fictitious example)
607
Millar
Table 1 Risk factors and risk factor weightings for clinical venous thromboembolism (VTE) in medical patients by multivariate analysis as reported
by Edelsberg et al.7 and Alikhan et al.8†
Risk factor on multivariate analysis
Relative risk
Edelsberg et al. Alikhan et al.
Female sex
Age > 75 years
Peripheral vascular disease
Obstructive airway disease
during admission
Cancer
Acute infectious disease
Heart failure during admission
Neurological disease with
paresis or paralysis
Post-thrombotic syndrome
Prior VTE
Intensive care unit admission or
procedure within previous 30 days
Operative procedure in previous 30 days
1.13
1.03
1.68
1.33
1.67
1.62
1.74
1.72
1.35
2.00
6.14
1.35
2.06
1.81
†
Values in the former, expressed as a hazard ratio, were determined in
92 162 patients admitted to managed care facilities in the USA on the
basis of reimbursement claims; the latter was based on an analysis of the
MEDENOX study4 data in 866 patients and expressed relative risk as an
odds ratio. Values specifying lower than average risk (‘acute coronary
syndromes’ and ‘other coronary heart disease during index admission’ in
study 1; ‘chronic respiratory disease’ in study 2) have been omitted.
should be excluded from thromboprophylaxis because
the B:H ratio would be less than the agreed value of R.
In the sensitivity analysis, the upper and lower 95%
confidence limits (CL) for baseline risk and RRR for each
study end-point were substituted for the measured
values in the analyses, individually or in combination.
Uncertainty over the major bleeding hazard was studied
using half or double the observed difference in bleeding
rates between the two groups. 95% confidence intervals
were not used for bleeding because the lower 95% CL
had a meaningless negative value and because the main
clinical interest is in patients who are at a higher than
average bleeding risk.
Results
Table 2 shows the RR required to achieve R-values from
3 to 10 for each of the three study end-points. For PE
alone, the estimated RR of PE in the groups of patients
who should receive prophylaxis, compared with the
PREVENT trial group, are 6.5, 10.8, 15.1 and 21.6,
respectively. With the other two end-points, lesser values
of RR satisfy the minimum agreed value of R (Table 2)
because of the higher incidence. However, the results
suggest that the PREVENT trial group is at too low a VTE
risk to satisfy any value of R in the range studied.
608
The conclusions from the primary data source were
strengthened by sensitivity analysis (Table 2). In univariate analysis, the results were most sensitive to an increase
in the baseline VTE rate and reduction in bleeding risk,
as expected. In multivariate analysis, RR values that
approximated unity (i.e. the minimum value of R was
satisfied by the PREVENT trial group) were found only
when all the sensitivity variables assumed values that
biased the result in the same direction.
The meta-analyses provided equivocal support for this
conclusion. Wein et al. provided results only for ‘PE
alone’, because symptomatic DVT rates were not reported
separately.9 Restricting the comparison to LMWH versus
no prophylaxis, RR values corresponding to above were
8.3, 13.9, 19.4 and 27.8 (Table 2). The other metaanalysis reported symptomatic DVT end-points and provided results more supportive of treating general medical
patients without further selection (Table 2).10
Discussion
The fundamental dictum in medical care is primum non
nocere. VTE prophylaxis using LMWH (or UH) carries a
small but finite risk of major bleeding that in effect sets
a lower limit of VTE risk to ensure that the dictum
is maintained, according to the minimum value of the
benefit–hazard ratio that is accepted as reasonable. This
idea has been investigated here by derivation of an equation relating the baseline risk of a VTE event and the
minimum accepted value of the ratio (R). The equation
has been derived in the setting of thromboprophylaxis,
but is capable of being applied to any disease in which a
balance obtains between a benefit expressed as a RRR
and a fixed adverse reaction risk. Some patients may be at
higher bleeding risk than the group value. This does not
invalidate the equation, which applies to a group, but
needs to be included in the decision to use prophylaxis.
For patients at high bleeding risk, the value of RR at the
chosen value of R underestimates the actual value in that
individual and militates against prophylaxis (Table 3).
The adverse event of interest is major bleeding. This
classification necessarily denotes an event of clinical
significance, although the precise definition may vary
between trials. In the PREVENT3 and MEDENOX4
studies, major bleeding was classified as intraocular,
spinal or epidural, intracranial or retroperitoneal bleeding or bleeding sufficient to cause a fall in haemoglobin of
at least 2 g/dL, transfusion of at least two units of blood,
or death. A practitioner making a judgement of the
minimum acceptable value of R needs to balance the
clinical significance of that degree of bleeding with
the significance of the disease being reduced by prophylaxis. In general, the method requires that R should be
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Benefit–hazard ratio in medical VTE prophylaxis
Table 2 Estimates of the relative risk (RR), for each of three symptomatic venous thromboembolism (VTE) end-points as shown, which are required to
ensure that the ratio of benefit–hazard R (VTE events avoided per major bleeding provoked by prophylaxis) is at least the value of R shown, in a group of
medical patients for whom prophylaxis against venous thrombosis is considered†
VTE end-point
VTE event rate
Major bleeding
Placebo
LMWH
Placebo
LMWH
0.0034
0.0028
0.0016
0.0049
PE or proximal
symptomatic DVT
0.0074
0.0039
0.0016
0.0049
PE or symptomatic DVT
0.0097
0.0056
0.0016
0.0049
Meta-analysis (Dentali et al.)
PE alone
0.0049
0.0020
0.0044
0.0058
PE or symptomatic DVT
0.0130
0.0058
0.0044
0.0058
Meta-analysis (Wein et al.)
PE alone
0.0099
0.0044
0.0109
0.0230
PREVENT study
PE alone
K
k’
R
A
RR
0.0033
0.006875
3
5
7
10
3
5
7
10
3
5
7
10
0.0206
0.0344
0.0481
0.0688
0.0206
0.0344
0.0481
0.0688
0.0206
0.0344
0.0481
0.0688
6.1
10.1
14.2
20.2
2.8
4.6
6.5
9.3
2.1
3.5
5.0
7.1
0.0014
0.0026
3
5
7
10
3
5
7
10
0.00792
0.01321
0.01849
0.02642
0.00792
0.01321
0.01849
0.02642
1.6
2.7
3.8
5.4
0.6
1.0
1.4
2.0
0.0121
0.0275
3
5
7
10
0.08239
0.13732
0.19224
0.27463
8.3
13.9
19.4
27.8
†
Values are obtained by application of Equation 2. The actual value of R is obtained by consensus or according to the judgement of the treating physician.
Data on symptomatic VTE and major bleeding rates are from the PREVENT study.2 Symbol k is the incremental rate of major bleeding found with
thromboprophylaxis; k’ is k divided by the relative risk reduction for each end-point shown in the trial, and A is the absolute risk of the respective VTE
end-point that satisfies the value assigned to R. RR is obtained by dividing A by the baseline (placebo group) trial-based risk of each VTE end-point,
respectively. DVT, deep venous thrombosis; PE, pulmonary embolism.
determined after defining which benefit is worthy of
prophylaxis and which bleeding risk is deemed to be
undesirable. For example, one may restrict the benefit to
one of the three end-points studied here and choose
between major bleeds or all bleeds. In other words, R
must be determined with due diligence, and in the full
knowledge of the prophylactic benefit and bleeding
hazard, rather than casually or intuitively. Note that if the
adverse event is discounted entirely the method fails
because the denominator in Equation 1 is then 0.
The clinical problem of which patients should receive
medical prophylaxis resolves under the approach
described here to a consensus (or individual) judgement
on the minimum acceptable value of R with respect to the
chosen VTE event. The results suggest that if R = 3 then
prophylaxis is indicated in a patient whose RR is 6.5, 3.0
and 2.3 for the three end-points studied (PE alone, PE or
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
proximal DVT and PE or any DVT), compared with the
PREVENT study group. Higher baseline risk values are
required at higher values of R (Table 2). I have assumed
that a value of R < 3 would not be considered reasonable
in relation to major bleeding.
The reasons for using the PREVENT study have been
listed. Although it is the largest study of LMWH prophylaxis in medical patients to date, it was underpowered for
symptomatic events, and hence uncertainty attaches to
the PE and DVT rates. Two approaches were devised to
compensate. First, a sensitivity analysis was carried out at
the 95%CL for these event rates. The effect of varying
bleeding rates was also carried out. The results (Table 3)
support the initial findings, that safe thromboprophylaxis
can only be achieved in subpopulations at higher VTE
risk that was found in the PREVENT trial. Only in a
multivariate analysis in which all variables had values set
609
Millar
Table 3 Results of sensitivity analysis, showing the relative risk estimates
obtained using Equation 2 (see text) when values of venous thromboembolism and major bleeding risk are varied as described in the text, alone
or in combination, for each value of benefit–hazard ratio (R) in the range
3–10†
Factor
Relative risk
R
1
2
3
4
5
6
7
8
9
10
11
†
Baseline risk,
lower 95%CL
3
5
7
10
Baseline risk,
3
upper 95%CL
5
7
10
RRR, lower 95% CL
3
5
7
10
RRR, upper 95%CL
3
5
7
10
1+3
3
5
7
10
1+4
3
5
7
10
2+3
3
5
7
10
2+4
3
5
7
10
Bleeding risk halved
3
5
7
10
Bleeding risk doubled
3
5
7
10
2+4+9
3
5
7
10
PE alone PE or PSDVT PE or ASDVT
13.0
21.6
30.3
43.2
2.8
4.7
6.6
9.4
14.6
24.3
34.0
48.5
4.7
7.8
11.0
15.7
31.1
51.9
72.6
103.8
10.0
16.7
23.4
33.5
6.7
11.2
15.7
22.5
2.2
3.6
5.1
7.3
3.2
5.4
7.5
10.8
12.9
21.6
30.2
43.1
1.1
1.8
2.5
3.6
4.7
7.8
10.9
15.6
1.6
2.7
3.8
5.4
6.7
11.1
15.6
22.3
2.2
3.6
5.0
7.2
11.2
18.7
26.2
37.5
3.6
6.0
8.5
12.1
3.9
6.5
9.1
13.0
1.3
2.1
2.9
4.2
1.5
2.5
3.5
5.0
5.9
9.9
13.9
19.8
0.6
1.0
1.5
2.1
3.4
5.6
7.8
11.2
1.3
2.2
3.1
4.4
5.1
8.5
11.9
17.0
1.6
2.7
3.8
5.5
8.1
13.4
18.8
26.9
2.6
4.3
6.1
8.7
3.2
5.3
7.4
10.6
1.0
1.7
2.4
3.4
1.1
1.9
2.6
3.8
4.5
7.6
10.6
15.1
0.5
0.9
1.2
1.7
End-points studied were pulmonary embolism (PE) alone, PE or proximal
symptomatic DVT (PE or PSDVT), and PE or any symptomatic deep venous
thrombosis (PE or ASDVT). CL, confidence limits; RRR, relative risk
reduction.
610
at the lower 95%CL (or half the observed rate in the case
of major bleeding) was the trial group VTE risk of borderline acceptability for non-hazardous prophylaxis. This
combination of values is implausible. Hence one can conclude that at the R-values in the range chosen for study,
thromboprophylaxis in medical patients may do more
harm than good, on the basis of the PREVENT study data.
Sensitivity analysis does not overcome the possibility
that the PREVENT study results were affected by the play
of chance and that the symptomatic thrombotic endpoint rates are an underestimate of the true value.
For this reason, I also analysed data from two recent
meta-analyses, which reported absolute rates of PE and
thrombotic events. Only one of these studies reported
symptomatic DVT rates.10 Although these studies appear
similar, they contain significant differences that are
reflected in the results shown here. Dentali et al. included
trials of LMWH, unfractionated heparin and fondaparinux, but excluded trials in ischaemic stroke and trials
with fewer than 30 patients.10 Wein et al. considered a
wider range of studies, including smaller studies, letters
or abstracts and stratified the data for each comparison of
interest.9 In their analysis of 11 trials of LMWH versus
placebo, five were in patients with ischaemic stroke. One
was a study of treatment rather than prophylaxis. The
baseline and incremental bleeding rates reported by the
two meta-analyses differ by factors of 8 and 12, respectively (Table 2) perhaps because of the inclusion of stroke
trials in the Wein et al. study. This causes a significant
difference in RR estimates in the present study and limits
the usefulness of the meta-analyses as supporting data.
However, Dentali et al. provide some evidence that the
trial-based group may be at a sufficiently high risk to
justify thromboprophylaxis.10 The data of Wein et al.
suggest that patients with stroke (possibly confounded by
age) are at a high bleeding risk and, on the basis of a
benefit–hazard analysis, should not receive thromboprophylaxis. In general, the heterogeneity of the metaanalyses limits application of the results to the question of
the value of thromboprophylaxis, without resorting to
the individual contributing trials.
The final issue is whether subpopulations at sufficiently
high risk to ensure acceptable R-values can be identified,
if the trial-based conclusions are correct. This is equivalent to asking whether additional risk factors for VTE can
be identified, which, if present in individual patients, can
be deemed to increase the VTE risk to the level required
to satisfy the chosen value of R and produce overall
benefit. For the end-point of highest incidence and hence
lowest RR at each R-value (PE or any DVT) the RR needs
to be in the range 2.3–7.6 for R ⱖ3 and ⱕ10. Table 1,
derived from the MEDENOX study8 and Edelsberg et al.7
(a retrospective study of hospital morbidity data in
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Benefit–hazard ratio in medical VTE prophylaxis
92 000 patients) shows the known risk factors that will
probably confer such a risk, alone or in combination.
Different weights for these and other risk factors measured in community studies may also be relevant. For
example, Heit et al. reported a relative multivariate risk
with recent surgery of 21.7, ‘trauma’ as 12.7, malignancy
with and without chemotherapy as 6.5 and 4.0, respectively, and varicose veins in patients less than 40 years of
age as 4.2.11 Thus, one issue to be resolved is which risk
factors should be included in the clinical assessment for
VTE prophylaxis and measurement of their relative
weights. My general conclusion is that several of the risk
factors shown in Table 3 have sufficient additional weight
to increase the VTE risk to a level at which R is satisfied,
especially at lower values of R.
The end-points studied here reflect emphasis of the
published work on PE, as the event most worthy of
preventing in thromboprophylaxis, but include DVT in
recognition of the immediate morbidity, probable effect
on hospital length of stay and possible chronic sequelae
such as post-phlebitic syndrome. However, the incidence
and clinical significance of chronec sequelae of DVT have
not been measured in medical patients. Thus, full application of the technique described here will require discussion of the relative clinical importance, in quantitative
terms, of uncomplicated symptomatic DVT versus fatal or
non-fatal PE and their prophylaxis.
The primary end-point in the PREVENT study and the
major bleeding hazard were measured at 21 days.
Ninety-day data were reported for the thrombotic endpoints, but not for bleeding, so estimation of the longterm data corresponding to the 21-day data is not
possible. If the bleeding hazard at 21 days is taken as a
proxy for the 90-day value, the RR, for R from 3 to 10,
vary from 2.5 to 8.3. This indicative result suggests that
the results of this study apply at time points other than
the 21-day interval specified in PREVENT. For the purposes of this study, hazard was restricted to major bleeding because only this is similar in clinical significance to
PE and may be life threatening. LMWH injections are
uncomfortable and cause unsightly local bruising but
these are relatively minor (but patient-unfriendly) disadvantages. For similar reasons, bleeding not amounting to
major bleeding seems inadequate as a measure of hazard
compared with the risk of PE. HITS is rare when using
LMWH.
Overall, the results of this study suggest that only a
small proportion of medical patients have sufficiently
strong risk factors for thrombosis for non-hazardous VTE
prophylaxis using LMWH at values of benefit–hazard
ratio probably regarded as reasonable by physicians in
medical patients. However, some opposing evidence was
found in one of two meta-analyses. The data from meta© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
analysis suggest that the overall conclusion applies
especially to patients with ischaemic stroke, whose
inclusion in current Australian guidelines is therefore
questionable.12
References
1 The National Institute for Clinical Studies. NICS Venous
Thromboembolism Prevention Program: Project planning
template (2005) [cited 2008 Sep]. Available from URL:
http://www.nhmrc.gov.au/nics/data/mediacache/
21036001178771897078/008%20NICS%20VTE%
20Prevention%20Program%20project%20plan%
20template%20%5BPDF%20276KB%5D.doc
2 Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR,
Colwell CW et al. Prevention of venous
thromboembolism. Seventh ACCP Conference on
Antithrombotic and Thrombolytic Therapy. Chest 2004;
126: 338S–400S.
3 Leizorovicz A, Cohen AT, Turpie AGG, Olsson C.-G,
Vaitkus PT, Goldhaber SZ. PREVENT Medical
Thromboprophylaxis Study Group. Randomized,
placebo-controlled trial of dalteparin for the prevention
of venous thromboembolism in acutely ill medical
patients. Circulation 2004; 110: 874–9.
4 Samama MM, Cohen AT, Darmon J-Y, Desjardins L,
Eldor A, Janbon C et al. Prophylaxis in Medical Patients
with Enoxaparin Study Group. A comparison of
enoxaparin with placebo for the prevention of venous
thromboembolism in acutely ill medical patients. N Engl
J Med 1999; 341: 793–800.
5 Shorr AF, Jackson WL, Weiss BM, Moores LK.
Low-molecular weight heparin for deep vein thrombosis
prophylaxis in hospitalised medical patients: results from
a cost-effectiveness analysis. Blood Coagul Fibrinolysis
2007; 18: 309–16.
6 McGarry LJ, Thompson D, Weisntein MC, Goldhaber SZ.
Cost effectiveness of thromboprophylaxis with a low
molecular weight heparin versus unfractionated heparin
in acutely ill medical patients. Am J Manag Care 2004; 10:
632–42.
7 Edelsberg J, Hagiwara M, Taneja C, Oster G. Risk of
venous thromboembolism in hospitalised medically ill
patients. Am J Health-Syst Pharm 2006; 63(Suppl. 6):
S16–21.
8 Alikhan R, Cohen AT, Combe S, Samama MM,
Desjardins L, Eldor A et al. Risk factors for venous
thromboembolism in hospitalised patients with acute
medical illness: analysis of the MEDENOX study. Arch
Intern Med 2004; 164: 963–8.
9 Wein L, Wein S, Haas S.-J, Shaw J, Krum H.
Pharmacological venous thromboembolism prophylaxis
in hospitalized medical patients. Arch Intern Med 2007;
167: 1476–86.
611
Millar
10 Dentali F, Douketis JD, Gianni M, Lim W, Crowther MA.
Meta-analysis: anticoagulant prophylaxis to prevent
symptomatic venous thromboembolism in hospitalized
medical patients. Ann Intern Med 2007; 146: 278–
88.
11 Heit JA, Silverstein MD, Mohr DN, Petterson TM,
O’Fallon WM, Melton J. Risk factors for deep vein
thrombosis and pulmonary embolism: a
612
population-based case-control study. Arch Intern Med
2000; 160: 809–15.
12 The Australia and New Zealand Working Party on the
Management and Prevention of Venous
Thromboembolism. Prevention of Venous
Thromboembolism 4th edition. Best Practice Guidelines for
Australia and New Zealand, Health Education &
Management Innovations Pty Ltd, 2007.
© 2009 The Author
Journal compilation © 2009 Royal Australasian College of Physicians
Internal Medicine Journal 39 (2009) 613–623
B R I E F C O M M U N I C AT I O N S
Malignant fibrous histiocytoma complicating nephrogenic
systemic fibrosis post liver transplantation
imj_1977
613..623
K. So,1 G. C. MacQuillan,1,2 L. A. Adams,1,2 L. Delriviere,1 A. Mitchell,1 H. Moody,3 D. J. Wood,2,4
R. C. Junckerstorff5 and G. P. Jeffrey1,2
1
Western Australian Liver Transplantation Service, 3Department of Nephrology, Sir Charles Gairdner Hospital, 4Perth Orthopaedic Institute, 5Division of
Tissue Pathology, PathWest Laboratory Medicine WA, QEII Medical Centre, Nedlands, and 2School of Medicine and Pharmacology, Faculty of Medicine
and Dentistry, The University of Western Australia, Perth, Western Australia, Australia
Key words
liver transplantation, systemic fibrosis,
histiocytoma.
Correspondence
Kenji So, Western Australian Liver Transplant
Service, 6th Floor, Sir Charles Gairdner
Hospital, Verdun Street, Nedlands, WA 6009,
Australia.
Email: [email protected]
Abstract
A 46-year-old man with cirrhosis secondary to hepatitis C virus infection and
alcohol underwent orthotopic liver transplantation, which required urgent
re-grafting because of biliary sepsis from necrosis of the left liver lobe. Recovery was complicated by renal failure and nephrogenic systemic fibrosis (probably related to intravenous gadolinium exposure). He subsequently developed
a malignant fibrous histiocytoma. We present this case highlighting the occurrence of two rare conditions in the same patient following liver transplantation. We believe this is the first case of its kind to be reported.
Received 31 May 2008; accepted 23
September 2008.
doi:10.1111/j.1445-5994.2009.01977.x
A 46-year-old man underwent orthotopic liver transplantation for cirrhosis secondary to hepatitis C virus infection
and alcoholic liver disease. His liver disease had been
complicated by hepatic encephalopathy, ascites, oesophageal varices, anaemia and spontaneous bacterial peritonitis. He was previously a non-responder to antiviral
treatment with interferon and ribavirin. His past medical
history included a repaired umbilical hernia and pancreatitis. He had normal renal function preoperatively.
The postoperative course was complicated by oliguria,
a rising serum creatinine level (peaking at 640 mmol/L
(normal 60–110 mmol/L), with a glomerular filtration
rate of 14 mL/min (normal >60 mL/min). He commenced haemodialysis through a femoral vascular
catheter on Day 2. He spent a total of 9 days in the
intensive care unit, and continued to be dialysis dependent. On the ward, he had worsening liver function
Funding: None.
Conflict of interest: None.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
tests, with the bilirubin peaking at 496 mmol/L (normal
<20 mmol/L). A magnetic resonance imaging (MRI) scan
of the patient’s liver with gadodiamide (14 mL Omniscan,
GE Healthcare, Australia) contrast material on Day 12
showed significant intrahepatic biliary duct dilatation,
and an abdominal computed tomography (CT) scan over
a week later showed extensive areas of ischaemia and
necrosis of the left liver lobe. This was associated with
biliary sepsis and he was subsequently re-listed for a
second liver transplant, which he received 22 days after
the first. He continued intermittent haemodialysis for
3 weeks after his second transplantation, by which time
his renal function had improved to baseline.
Four days before the second transplant, he developed
sudden-onset of intermittent deep limb pain (upper limbs
more affected than the lower limbs), which was also
extremely sensitive to any light touch. His lower limbs
became very oedematous. He developed bilateral lower
limb weakness and paraesthesia in his finger tips and
toes. On day 19 post second transplant an erythematous,
raised, confluent and blanching rash appeared on his
613
So et al.
Figure 1 Severely oedematous and indurated lower limbs.
Figure 2 The large right anterolateral hip mass following excisional
biopsy.
lower abdomen, which was not painful or pruritic. Over
the next week, the skin changes progressed, with more
notable induration of the skin on the thighs, abdomen
and arms. A skin biopsy from the lower abdomen showed
increased dermal mucin, increased dermal cellularity
with proliferation of fibroblasts and thickening of
collagen fibres, all of which were consistent with
nephrogenic systemic fibrosis (NSF).1 He commenced
daily plasmapheresis, and penicillamine was started
10 days later.
Plasmapheresis was ceased after 2 weeks without any
obvious improvement. Because of progressive skin
disease despite ongoing intensive physiotherapy, he
received high-dose intravenous steroids (1 g methylprednisolone once a day) for 3 days and then commenced
intravenous immunoglobulin infusions (five doses).1
There was some transient improvement in his upper limb
movements, but overall, the fibrosis progressed. Four
further courses of immunoglobulin were administered
and thalidomide was commenced some two and a half
months post NSF diagnosis.1 Over the next few months,
his upper limb function improved, but his lower limbs
remained severely oedematous with markedly reduced
movements and a fixed plantar flexion (Fig. 1) such
that he was wheelchair dependent and unable to leave
hospital.
The patient was found to have a right anterolateral hip
mass following a fall, 5 months after the initial development of the NSF. Clinically, it was thought to be an
organizing haematoma. It was extremely painful. An
initial attempt of aspiration under ultrasound guidance
confirmed the presence of blood. He underwent debridement on multiple occasions and had several courses of
intravenous antibiotics (for presumed infection, with a
raised C-reactive protein and white cell count). It is
unclear whether the hip mass was already present before
the fall. It continued to enlarge over a 3- to 4-month
period. An excision biopsy was performed (Figs 2,3). Histopathological examination revealed a pleomorphic cellular neoplasm composed of epithelioid and spindled cells
involving the dermis, subcutis and focally the skeletal
muscle and iliac bone (Fig. 4). The neoplastic cells exhibited marked nuclear pleomorphism and up to 40 mitotic
figures per 10 high power fields were identified. Focal
lymphovascular space invasion was present. Immunohistological staining for broad spectrum cytokeratin was
negative (AE1/AE3 and MNF116), excluding epithelial
differentiation. The melanoma markers S-100, HMB45
and Melan-A were negative. Muscle differentiation was
excluded with negative results on staining for smooth
muscle actin, calponin, caldesmon, desmin, myogenin
and Myo-D1. A diagnosis of undifferentiated high grade
pleomorphic sarcoma (malignant fibrous histiocytoma,
614
Figure 3 Excisional biopsy of the hip mass (macroscopic view).
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
MFH complicating NSF
Figure 4 Pleomorphic cellular neoplasm composed of epithelioid and
spindled cells involving the dermis; subcutis and focally the skeletal
muscle and iliac bone.
MFH) was made. Staging CT scans of his chest demonstrated pulmonary metastases and he underwent
palliative resection of the fungating pelvis mass. He subsequently died 14 months after the development of the
diagnosis of NSF.
NSF, a fibrosing skin disorder that occurs almost exclusively in patients with renal failure, was first recognized
in 1997 by Cowper et al.2 It is characterized by thickening
and hardening of the skin overlying the trunk and
extremities. More recently, there has been evidence that
NSF involved other organs, including skeletal muscles,
diaphragm, pleura, pericardium, myocardium, dura
mater and vessels of the heart.3–7 There have been over
200 cases of NSF reported worldwide including this case,
previously published by Caccetta and Chan.1 The
subsequent presentation of MFH following a diagnosis of
NSF has not been described previously.
NSF occurs equally in both sexes, ranging in age from
8 to 87 years old. The primary cutaneous lesions appear
in the lower extremities in the majority of patients
(distribution from ankles to mid-thighs), followed by the
upper extremities (from wrists to mid-upper arms) and
trunk.3,6 It often presents initially with oedema, which
gradually resolves, leaving erythematous indurated
plaques, which can progress to thickened and hardening
of the skin with brawny pigmentation, peau d’orange
appearance and formation of papules and subcutaneous
nodules.1,4,6,7 The face is rarely involved.7,8 Most patients
describe the lesions as painful or pruritic, and if joints
are involved, it can lead to contractures resulting in a
reduced level of functioning.3,9 Calciphylaxis has also
been reported in association with NSF.3,10
Apart from renal impairment, other comorbidities have
been reported to be associated with NSF. A subset of
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
individuals with NSF had surgical procedures preceding
its onset (mainly surgery with a vascular reconstructive
component).1,7,11 There have been no known reports of
MFH in patients with NSF.
There are no specific laboratory tests to aid the diagnosis of NSF. A deep skin biopsy is essential for definitive
diagnosis. Lesions contain a proliferation of dermal
spindle cells with long dendritic processes, which may
extend into the subcutaneous tissue. Immunohistochemical staining is positive for CD34 and procollagen.
There may also be positive staining for CD68 and factor
XIIIa. Inflammatory cells are usually absent.3–6,11
There is increasing evidence of a connection between
gadolinium-containing contrast agent and the occurrence
of NSF. Gadolinium is a lanthanide ion that is excreted
almost exclusively by the kidneys.12 Grobner13 reported
five patients with end-stage renal disease on haemodialysis, who developed NSF 2–4 weeks following
undergoing magnetic resonance angiography with
administration of gadolinium-containing contrast.
Marckmann et al. also found a correlation between the
occurrence of NSF and an exposure to gadodiamide
25 days earlier, in 13 patients with end-stage renal failure
on haemodialysis, but without involvement of a coexistent acidosis.14 In a pilot study, gadolinium was detected in
the tissue of a number of patients with NSF, supporting
an epidemiological association between exposure to
gadolinium-containing contrast material and development of NSF.15 Of note, our patient underwent an MRI
with gadolinium 30 days prior to his diagnosis of NSF.
More extensive recent studies have supported the
ever-increasing evidence for the relationship between
gadolinium and NSF, most significant after gadodiamide
than any other gadolinium-based agents.16–18
At present, there is no proven therapy for NSF. Numerous treatment options have been explored with inconsistent results. There have been reports of success with
plasmapheresis for treating NSF, and in particular, in
three patients following liver transplantation.8 A study by
Gilliet et al. found extracorporeal photopheresis therapy
to be successful in softening the skin lesions and improving joint mobility.19 It is likely that an improvement in
renal function can halt or reverse the process, which
raises the possibility of renal transplantation as a treatment option.
As far as we are aware, there have been no cases
reported of MFH or undifferentiated pleomorphic
sarcoma arising in a patient with NSF. Literature search
has found no reported link between the two. Both NSF
and MSH appear to be mesenchymal in origin.20 At the
Perth Bone Tumour Registry there have been three cases
of MFH associated with fibrosis, one with a metallic
implant, one with a total hip replacement and one with
615
So et al.
an area of osteomyelitis following shrapnel wound
sustained in the Second World War.21
MFH is thought to occur de novo (i.e. not from a
pre-existing benign lesion). No defined aetiology has
been identified, but several associated or predisposing
factors have been reported, including genetic predisposition, exposure to radiation or chemotherapy,
chemical carcinogens, chronic irritation and lymphoedema.22 They present with a mass and are more commonly painless. However, pain may be present following
an injury, and may wax and wane over time. Systemic
symptoms, such as fever, weight loss and malaise, are
generally absent.
Laboratory evaluation is usually normal, except for
elevations in alkaline phosphatase (in approximately
40%), lactate dehydrogenase (in approximately 30%)
and erythrocyte sedimentation rate. Laboratory abnormalities do not correlate with disease extent. There are
many different subtypes and histopathological variants,
and MFH is essentially a diagnosis that is usually made
following the exclusion of other lines of cell differentiation by the use of a panel of immunohistological markers.
Adequate tissue biopsy and then examination by an
experienced histopathologist, using immunohistochemistry, is essential.23 For localized tumours, surgical resection is the only curative option, with or without adjuvant
radiotherapy or chemotherapy. Definitive radiotherapy is
used when surgery is not an option.24
We have presented a case of a patient receiving liver
transplantation, which has been complicated by NSF and
undifferentiated sarcoma. We believe this is the first case
of both occurring in the one patient. Further research is
needed to determine if in fact there is an association
between the two.
References
1 Caccetta T, Chan J. Nephrogenic systemic fibrosis
associated with liver transplantation, renal failure and
gadolinium. Australas J Dermatol 2008; 49: 48–51.
2 Cowper SE, Robin HS, Steinberg SM, Su LD, Gupta S,
LeBoit PE. Scleromyxedema-like cutaneous disease in
renal-dialysis patients. Lancet 2000; 356: 1000–1.
3 Galan A, Cowper SE, Bucala R. Nephrogenic systemic
fibrosis (nephrogenic fibrosing dermopathy). Curr Opin
Rheumatol 2006; 18: 614–17.
4 Weiss AS, Scott Lucia M, Teitelbaum I. A case of
nephrogenic fibrosing dermopathy/nephrogenic systemic
fibrosis. Nat Clin Pract Neurol 2007; 3: 111–15.
5 Gibson SE, Farver CF, Prayson RA. Multiorgan
involvement in nephrogenic fibrosing dermopathy. Arch
Pathol Lab Med 2006; 130: 209–12.
6 Mendoza FA, Artlett CM, Sandorfi N, Latinis K,
Piera-Velazquez S, Jimenez SA. Description of 12 cases of
616
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
nephrogenic fibrosing dermopathy and review of the
literature. Semin Arthritis Rheum 2006; 35: 238–49.
Kucher C, Steere J, Elenitsas R, Siegel DL, Xu X.
Nephrogenic fibrosing dermopathy/nephrogenic systemic
fibrosis with diaphragm involvement in a patient with
respiratory failure. J Am Acad Dermatol 2006; 54: S31–4.
Baron PW, Cantos K, Hillebrand DJ, Hu K, Ojogho ON,
Nehlsen-Cannarella S et al. Nephrogenic fibrosing
dermopathy after liver transplantation successfully
treated with plasmapheresis. Am J Dermatopathol 2003;
25: 204–9.
Boyd AS, Zic JA, Abraham JL. Gadolinium deposition in
nephrogenic fibrosing dermopathy. J Am Acad Dermatol
2007; 56: 27–30.
Edsall LC, English JC III, Patterson JW. Calciphylaxis and
metastatic calcification associated with nephrogenic
fibrosing dermopathy. J Cutan Pathol 2004; 31: 247–53.
Cowper SE. Nephrogenic fibrosing dermopathy: the first
6 years. Curr Opin Rheumatol 2003; 15: 785–90.
Endre ZH. Nephrogenic systemic fibrosis: is any contrast
safe in renal failure? Intern Med J 2007; 37: 429–31.
Grobner T. Gadolinium – a specific trigger for the
development of nephrogenic fibrosing dermopathy and
nephrogenic systemic fibrosis? Nephrol Dial Transplant
2006; 21: 1104–8.
Marckmann P, Slov L, Rossen K, Dupont A,
Damholt MB, Heaf JG et al. Nephrogenic systemic
fibrosis: suspected causative role of gadodiamide used for
contrast-enhanced magnetic resonance imaging. J Am Soc
Nephrol 2006; 17: 2359–62.
High WA, Ayers RA, Chandler J, Zito G, Cowper SE.
Gadolinium is detectable within the tissue of patients
with nephrogenic systemic fibrosis. J Am Acad Dermatol
2007; 56: 21–6.
Thomsen HS, Morcos SK. Nephrogenic systemic fibrosis:
more questions and some answers. Nephron Clin Pract
2008; 110: c24–32.
Wertman R, Altun E, Martin DR, Mitchell DG,
Leyendecker JR, O’Malley RB et al. Risk of nephrogenic
systemic fibrosis: evaluation of gadolinium chelate
contrast agents at four American universities. Radiol
2008; 248: 799–806.
Prince MR, Zhang H, Morris M, MacGregor JL,
Grossman ME, Silberzweig J et al. Incidence of
nephrogenic systemic fibrosis at two large medical
centres. Radiol 2008; 248: 807–16.
Gilliet M, Cozzio A, Burg G, Nestle FO. Successful
treatment of three cases of nephrogenic fibrosing
dermopathy with extracorporeal photopheresis. Br J
Dermatol 2005; 152: 531–6.
Dei Tos AP. Classification of pleomorphic sarcomas:
where are we now? Histopathology 2006; 48:
51–62.
Bone Tumour Registry of Western Australia, Perth
Orthopaedic Institute, UWA, Nedlands, WA.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
MFH complicating NSF
22 Zahm SH, Fraumeni JF Jr The epidemiology of soft tissue
sarcoma. Semin Oncol 1997; 24: 504–14.
23 Heslin MJ, Lewis JJ, Woodruff JM, Brennan MF. Core
needle biopsy for diagnosis of extremity soft tissue
sarcoma. Ann Surg Oncol 1997; 4: 425.
24 Kepka L, Suit HD, Goldberg SI, Rosenberg AE,
Gebhardt MC, Hornicek FJ et al. Results of radiation
therapy performed after unplanned surgery (without
re-excision) for soft tissue sarcomas. J Surg Oncol 2005;
92: 39.
A case of uterine tumour resembling ovarian sex cord tumour
responding to second-line, single agent anastrazole
P. Blinman1 and M. H. N. Tattersall1,2
1
Sydney Cancer Centre, Sydney, NSW, Australia, 2Department of Medicine, University of Sydney, Sydney, NSW, Australia
Key words
endometrial stromal tumour, UTROSCT,
aromatase inhibitors, anastrazole.
Correspondence
Prunella Blinman, NHMRC Clinical Trials Centre,
Looked Bag 77, Camperdown, NSW 1450,
Australia.
Email: [email protected]
Abstract
Uterine tumour resembling ovarian sex cord tumour (UTROSCT) are a histological variant of endometrial stromal sarcomas (ESS). There is no established
medical management of metastatic UTROSCT or ESS, although there is evidence supporting the use of hormonal therapy. Given the success of aromatase
inhibitors in breast cancer, their potential role in ESS and UTROSCT is of
current interest. We report the first case of response to second-line, single
agent anastrazole in a patient with metastatic UTROSCT.
Received 24 March 2008; accepted 7 July 2008.
doi:10.1111/j.1445-5994.2009.01998.x
A 49-year-old female non-smoker presented in 1988
with right iliac fossa pain. A pelvic ultrasound suggested
a uterine fibroid for which she had a hysterectomy.
Macroscopically, the uterus had a smooth endometrium
with a 6.5-cm myometrial mass in the left lateral wall.
Microscopically, the mass showed anastamosing trabeculae of small round to oval epithelial-like cells which
appeared to form small glands. Immunohistochemical
staining was negative for epithelial membrane antigen,
diffusely positive for vimentin, CD 99 and alpha inhibin
with some positive staining for desmin. The histopathological diagnosis was low-grade endometrial stromal
sarcoma (ESS) of the sub-type uterine tumour resembling ovarian sex cord tumour (UTROSCT). Pelvic
radiotherapy was given due to the closest margin of
0.5 mm.
The patient remained well until 1999 before presenting
with upper abdominal firmness. A computed tomography (CT) scan showed an 8-cm mass extending from the
left renal vein to the left common iliac vein which almost
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
surrounded the aorta at the level of the L3 vertebral body.
There was left hydronephrosis and a filling defect in the
inferior vena cava (IVC) suggestive of tumour thrombus.
She was commenced on tamoxifen, warfarin and given
two cycles of neoadjuvant doxorubicin with minimal
response. En bloc resection of the tumour, left kidney and
infrarenal aorta and removal of the left iliac and renal
veins along with an axillo-bifemoral bypass was performed. The histopathology confirmed recurrent ESS
involving the surrounding small vessels but not the
kidney, ureter or aorta.
Annual CT scans for the next 4 years showed stable
tumour IVC thrombus only. However, in 2004 the patient
developed dyspnoea and a CT scan revealed a large caval
thrombus extending through the right atrium into the
right pulmonary arteries, and pulmonary infarcts in the
right lower lobe. Positron emission tomography scanning
showed 18-fluoro-deoxyglucose avid lesions along the
entire course of the IVC in addition to left para-aortic
and common iliac nodal regions (Fig. 1). A transthoracic
617
Blinman & Tattersall
Figure 1 Positron emission tomography scan in 2004 showing extensive
18-fluoro-deoxyglucose avid tumour thrombus from the inferior vena
cava to the right atrium.
echocardiogram demonstrated a large mass in the IVC
and right atrium which entered the right ventricle during
systole. Tamoxifen was changed to anastrazole 1 mg daily
and home oxygen was arranged for symptomatic relief.
A progress CT scan performed in 2005 showed a
marked reduction in the size of the tumour thrombus in
the right atrium and pulmonary arteries and partial resolution of the extensive IVC thrombus (Figs 2,3). The
patient was last seen in December, 2007 at which time
home oxygen was no longer required. A repeat CT scan
showed further improvement with reduction in the
lymphadenopathy.
Endometrial stromal sarcoma are rare tumours that
account for 7–15% of uterine sarcomas.1 The annual
incidence of ESS is 1–2 per million women with less than
700 new cases each year in Europe.2 They have a
typically indolent growth pattern and may relapse many
years after surgery3 with a mean time to recurrence
6.8 years.4 UTROSCT are considered to be a rare histological variant of ESS.
The role of aromatase inhibitors in locally recurrent
and metastatic UTROSCT is not clearly established. Letro-
Figure 2 Contrast computed tomography scan showing thrombus
extending from the inferior vena cava into the right atrium in 2004 (left)
with partial resolution of the thrombus in 2005 (right).
618
Figure 3 Contrast computed tomography scan showing extensive
thrombus in the right pulmonary artery in 2004 (left) with substantial
improvement in 2005 (right).
zole has been used in metastatic ESS as first-line therapy2
or second-line therapy after failure of progestins.5 Anastrazole has been used concurrently with megestrol
acetate in a patient with recurrent UTROSCT.6 The
present case is the first documented response to single
agent, and second-line, anastrazole in UTROSCT.
Patients with metastatic breast cancer who acquire
resistance to tamoxifen after initial response do not lose
expression of oestrogen receptors7 and 15–20% of such
patients will respond to second-line anastrazole or fulvestrant.8 Therefore, a possible mechanism of response to
second-line aromatase inhibitors is reduction of oestrogen production in peripheral tissues and consequent
growth inhibition of oestrogen receptor-dependent
cancers. We propose a similar mechanism of action for
second-line anastrazole in our patient, although the
‘response’ to tamoxifen is only circumstantial inferred
from the 4 years of stable disease while on tamoxifen
before progression was documented in 2004.
References
1 Acharya S, Hensley ML, Montag AC, Fleming GF. Rare
uterine cancers. [Erratum appears in Lancet Oncol. 2006
Feb; 7(2): 105]. Lancet Oncol 2005; 6: 961–71.
2 Pink D, Lindner T, Mrozek A, Kretzschmar A,
Thuss-Patience PC, Dorken B. et al. Harm or benefit of
hormonal treatment in metastatic low-grade endometrial
stromal sarcoma: single center experience with 10 cases
and review of the literature. [See comment]. Gynecol
Oncol 2006; 101: 464–9.
3 Styron SL, Burke TW, Linville WK. Low-grade
endometrial stromal sarcoma recurring over three
decades. Gynecol Oncol 1989; 35: 275–8.
4 Yilmaz A, Rush DS, Soslow RA. Endometrial stromal
sarcomas with unusual histologic features: a report of 24
primary and metastatic tumors emphasizing fibroblastic
and smooth muscle differentiation. Am J Surg Pathol
2002; 26: 1142–50.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
UTROSCT responding to second-line, single agent anastrazole
5 Maluf FC, Sabbatini P, Schwartz L, Xia J, Aghajanian C.
Endometrial stromal sarcoma: objective response to
letrozole. Gynecol Oncol 2001; 82: 384–8..
6 Leiser AL, Hamid AM, Blanchard R. Recurrence of
prolactin-producing endometrial stromal sarcoma with
sex-cord stromal component treated with progestin and
aromatase inhibitor. Gynecol Oncol 2004; 94: 567–71.
7 Johnston SRD, Saccani-Jotti G, Smith IE, Salter J,
Newby J, Coppen M et al. Changes in estrogen receptor,
progesterone receptor, and pS2 expression in
tamoxifen-resistant human breast cancer. Cancer Res
1995; 55: 3331–38.
8 Howell A, Robertson JFR, Quaresma Albano J,
Aschermannova A, Mauriac L, Kleeberg UR et al.
Fulvestrant, formerly ICI 182,780, is as effective as
anastrozole in postmenopausal women with advanced
breast cancer progressing after prior endocrine treatment.
J Clin Oncol. 2002; 20: 3396–403.
Fluoroquinolone-induced immune thrombocytopenia: a report
and review
imj_1996
619..629
1
C. Y. Cheah, B. De Keulenaer2 and M. F. Leahy3
1
Department of Haematology, 2Intensive Care Unit, 3Haematology Department, Fremantle Hospital, Fremantle, Western Australia, Australia
Key words
fluoroquinolones, ciprofloxacin, platelet
antigens, chemical-induced thrombocytopenia.
Correspondence
Chan Cheah, Fremantle Hospital, Alma Street,
Fremantle 6160 Western Australia, Australia.
Email: [email protected]
Received 21 February 2008; accepted 4 June
2008.
doi:10.1111/j.1445-5994.2009.01996.x
Abstract
Fluoroquinolones are an emerging but underrecognized cause of druginduced thrombocytopenia. Due to their broad spectrum they are often used
in empirical treatment of febrile neutropenic, thrombocytopenic patients following myelosuppressive chemotherapy. They are associated with a range
of immunohaematopathology. A 76-year-old male developed severe thrombocytopenia following treatment with ciprofloxacin on two occasions for
community-acquired pneumonia. The temporal association, response to
dechallenge, dramatic response to rechallenge and exclusion of other causes
combined with detection of platelet-reactive antibodies of the immunoglobulin G class against glycoprotein IIb/IIIa following ciprofloxacin rechallenge
makes causality probable. We present a brief review of immunohaematopathology associated with fluoroquinolones and draw attention to the structural
similarity between quinolones and quinine to explore potential mechanisms
for the phenomenon. Fluoroquinolones can induce drug-dependent, plateletreactive antibodies causing complement-mediated destruction of platelets. The
underlying mechanism to explain this is unclear; however, we hypothesize
that the chemical similarities shared with quinine may be contributory. When
using these agents clinicians should be aware of the possibility of drug-induced
thrombocytopenia or thrombotic thrombocytopenic purpura.
A wide variety of drugs has been observed to cause
thrombocytopenia by immunological mechanisms.
Recent reviews have focused on the incidence, causes
Funding: None.
Conflict of interest: None.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
and pathogenesis of drug-induced immune thrombocytopenia (DITP).1,2 Antibiotics are commonly responsible;
better known examples include rifampicin, sulfonamides,
linezolid and vancomycin.2 Fluoroquinolones are synthetic antibiotics widely used due to their broad spectrum, high oral bioavailability, tissue concentrations
and safety profile. They share a central quinolone or
619
Cheah et al.
naphthyridone nucleus, but different structural modifications have been made to alter the potency, pharmacokinetics and toxicity.3 Some drugs have been withdrawn due
to an unacceptably high rate of adverse events, notably
hepatotoxicity with trovafloxacin4 and haemolytic
uraemic syndrome (HUS) with temafloxacin.5 There is a
growing number of reports of fluoroquinolone-induced
thrombocytopenia and thrombotic thrombocytopenic
purpura (TTP). Comprehensive reviews of DITP2,6,7 and
TTP8 have not mentioned fluoroquinolones a potential cause. We present a case of ciprofloxacin-induced
immune mediated thrombocytopenia supported by the
localization of ciprofloxacin-induced platelet-reactive
antibodies and briefly review the literature on the topic.
Clinical case
A 76-year-old man with a history of emphysema,
hypercholesterolaemia and hypertension was admitted
to the intensive care unit with community-acquired
pneumonia and acute on chronic renal failure. Medications prior to presentation included pravastatin and
amlodipine, with no known adverse drug reactions. The
initial platelet count was 171 ¥ 109/L. Treatment consisted of intravenous ticarcillin/clavulanic acid 3.1 g
twice daily and ciprofloxacin 200 mg twice daily for
4 days. Unfractionated heparin, 5000 units twice daily
was given subcutaneously from day 1. Therapy was
switched to ceftriaxone on day 4 in response to microbiological susceptibilities. The platelet count progressively fell from 160 ¥ 109 to 120 ¥ 109/L during the
4 days of treatment but returned to 200 ¥ 109/L after
ciprofloxacin was ceased.
Clinical and radiological improvement followed, but
the patient remained ventilator dependent. On day 30
evidence of ventilator associated pneumonia (fever, with
new chest x-ray consolidation) developed. A single dose
of ciprofloxacin 400 mg was given intravenously. Within
12 h the platelet count had fallen from 285 ¥ 109 to
40 ¥ 109/L. Ciprofloxacin and heparin were ceased.
Platelet factor 4 enzyme-linked immunosorbent assay
(PF4 ELISA) for heparin-induced thrombocytopenia
and ciprofloxacin-induced platelet-reactive antibody test
were requested. Haemoglobin, leukocytes and coagulation profile remained within normal limits and there was
no evidence of haemolysis. Peripheral blood film showed
marked thrombocytopenia without microangiopathic
features. Petechiae, purpura and ecchymosis were absent.
No new focal neurological signs developed. No additional
treatment was implemented. The platelet count reached a
nadir of 20 ¥ 109/L on day 32 but improved rapidly,
reaching 172 ¥ 109/L by day 36. Repeat serum was drawn
at day 36 and platelet antibody testing was repeated.
Figure 1 illustrates the platelet count versus time.
PF4 ELISA for HITS was negative. However, repeat
testing on day 36 demonstrated reaction of patient serum
with patient and control platelets in the presence (but not
in the absence) of (native) ciprofloxacin added in vitro.
This finding supports the presence of ciprofloxacininduced platelet reactive antibodies. Monoclonal
antibody immobilization of platelet antibodies using
ciprofloxacin at a concentration of 1 mg/mL and monoclonal P2 antibodies identified Glycoproteins IIb/IIIa as
targets. Bone marrow aspiration was not performed due
to recovery, response to dechallenge, recurrence on
rechallenge and risk to patient. No other platelet
Figure 1 Graph of platelet count (y-axis,
¥109/L) versus time (x-axis, days) and
relationship to use of ciprofloxacin.
Course of ciprofloxacin induced
thrombocytopenia
350
Platelet count (x109/L)
300
250
200
150
100
50
0
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Time (days)
platelet count
ciprofloxacin
620
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Fluoroquinolone-induced immune thrombocytopenia
auto-antibodies were demonstrated. Heparin was recommenced at day 40 at the previous dose of 5000 units twice
daily without recurrent thrombocytopenia. The above
features, combined with demonstration of ciprofloxacin
dependent antibodies to platelet glycoprotein IIb/IIIa in
the patient’s serum satisfy all four criteria proposed by
George et al.6 to support a diagnosis of ciprofloxacininduced immune thrombocytopenia.
Discussion
We identified cases of thrombocytopenia, TTP and HUS
associated with fluoroquinolone by searching MEDLINE
and EMBASE to November 2007 using terms ‘thrombocytopenia’, ‘hemolytic uraemic syndrome’ or ‘thrombotic thrombocytopenic purpura’ with subheading
‘chemical induced’. This was combined with ‘fluoroquinolone’ or each generic name. The bibliographies were
used to identify additional reports. The Australian Therapeutic Goods Administration maintains a database of
adverse drug reactions (Adverse Drug Reports Advisory
Committee (ADRAC)) which was also searched.
In total 180 cases were identified. Detailed information was available in 17 articles9–25 (29 cases).
Cases considered ‘probable’ or ‘definite’ according to
published criteria6 were included. The drugs implicated were ciprofloxacin10,11,13,14,19,23 (eight cases),
pefloxacin12,17,18 (seven cases), nalidixic acid15 (six
cases), norfloxacin16,20,24 (three cases), temafloxacin9
(two cases), alatrofloxacin,21 levofloxacin25 and
tosufloxacin22 (one each).
Several patterns of quinolone-induced thrombocytopenia were identified. The first and most common
(23/29 cases) was isolated thrombocytopenia. The mean
platelet nadir was 24.9 ¥ 109/L reached after 9.6 days
(range 3–22), with recovery after 8 days (range 3–21).
7/23 had bone marrow aspirates, all of which showed
reactive marrow with increased megakaryocytic activity.
6/23 had platelet antibody testing performed and of
these 4/6 were positive, predominantly immunoglobulin G antibodies. Bleeding occurred in 10/23 of patients.
Ecchymoses or purpura were reported in six patients
followed by petechiae in four and epistaxis in three.
22/23 had the causative drug ceased, three patients
were treated with corticosteroids, two with platelet
transfusions and with intravenous immunoglobulin and
all patients survived. In one case halving the dose of
ciprofloxacin resulted in normalization of platelet count.
In a series of two patients treatment with ciprofloxacin was associated with aplastic anaemia.14 Both
patients were men in their 20s with indications for
treatment being pneumonia and typhoid fever. Each
received 5 days of ciprofloxacin, had previously normal
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
blood counts and no adverse drug reactions and subsequently developed severe pancytopenia. Bone marrow
aspirates in each case demonstrated marked hypocellularity suggestive of aplastic anaemia. Both developed
multiple bleeding complications and were managed
with transfusions of blood products with one surviving
and the other dying of unspecified bleeding complications. Although platelet antibody testing was not
performed, diagnoses of ciprofloxacin-induced bone
marrow failure were proposed.
Fluoroquinolones have been associated with TTP and
HUS. The largest is the description of ‘temafloxacin syndrome’ in which 95 patients developed varying degrees
of haemolytic anaemia, thrombocytopenia and oliguric
renal failure following treatment with temafloxacin,
which was withdrawn from market shortly afterwards.5
There are two reports of ciprofloxacin associated with
TTP. Gales and Sulak 21 describe a 53-year-old man who
developed microangiopathic haemolytic anaemia 3 days
after treatment with ciprofloxacin for empirical treatment of febrile neutropenia (29 days after induction
chemotherapy for acute lymphoblastic leukaemia). His
platelet count was 144 ¥ 109/L prior to commencement
of ciprofloxacin, and repeat bone marrow showed
response to chemotherapy. During treatment with ciprofloxacin the platelet count dropped to 40 ¥ 109/L and
oliguric renal failure (with microangiopathic blood film)
developed. Ciprofloxacin was ceased and treatment with
plasmapheresis and dialysis resulted in recovery of renal
function and platelet count and the patient went on to
achieve remission. Ciprofloxacin-dependent immunoglobulin G and immunoglobulin M antibodies were
detected. A second report by Mouraux and colleagues11
describes a 43-year-old previously healthy woman
treated with ciprofloxacin 500 mg daily for cystitis
presenting with petechiae, altered mental state, neck
stiffness and extensor plantar reflexes. Subsequent
investigation showed lacunar infarcts, subarachnoid
haemorrhage, renal failure, thrombocytopenia, microangiopathic blood film and sterile cerebrospinal fluid
without pleocytosis. Despite cessation of ciprofloxacin on
admission, platelet transfusion and plasmapheresis the
patient died.
Fluoroquinolones have been associated with a range of
immunohaematopathology, including isolated thrombocytopenia, TTP and bone marrow failure. The
mechanism of fluoroquinolone-induced immune thrombocytopenia might be explained by the structural relationship between fluoroquinolones and quinine, the first
described and best-studied example of drug dependent
antibody mediated thrombocytopenia.5,10,26 Quinine
differs in the addition of a long side chain at position 4
where quinolones have an oxygen molecule, as seen in
621
Cheah et al.
association, response to de-challenge and rechallenge,
and the finding of drug-dependent platelet reactive
antibodies strongly support the diagnosis. In isolated
thrombocytopenia drug withdrawal results in prompt
recovery of platelet count.
Acknowledgements
Ms Annette Hughes, Senior Scientist, Immunohaematology, Royal Perth Hospital, for helpful advice and performing flow cytometry and monoclonal antibody
immobilisation of platelet antigens. Ms Anna Allman,
Clinical Pharmacist, Intensive Care Unit, Fremantle
Hospital for searching ADRAC for case reports.
References
Figure 2 Chemical similarity between generic quinolone nucleus and
quinine.
Figure 2. Quinine is well known to cause plateletreactive antibody-mediated immune thrombocytopenia
and HUS-TTP.27–30 One proposed model to explain these
observations postulates that drugs such as quinine are
able to ‘improve the fit’ between weakly auto-reactive
antibodies complementarity determining regions and
target platelet glycoproteins, such as IIb/IIIa.31 The drug is
therefore able to facilitate higher affinity binding
between antibody and target by improving the structural
and chemical interaction. It is possible that this mechanism could also explain the action of fluoroquinolones
in this setting. Other authors have identified
fluoroquinolone-dependent antibodies against red blood
cells9,19 and platelets16,19,21; however, this report is the first
we are aware of localizing the binding site for
ciprofloxacin-induced platelet antibodies.
As ciprofloxacin is commonly used for empirical treatment of neutropenic sepsis in thrombocytopenic patients
following myelotoxic chemotherapy, clinicians should be
aware of the possibility of ciprofloxacin-induced thrombocytopenia or TTP as in this setting the diagnosis may be
overlooked and the event attributed to anti-neoplastic
agents.
Fluoroquinolone-induced immune thrombocytopenia
and TTP are rare but should be considered the differential
diagnosis of acute onset of thrombocytopenia after treatment with these drugs. There is no way to prospectively
identify which patients will be affected, but temporal
622
1 Von drygalski A, Curtis BR, Bougie DW, McFarland JG,
Ahl S, Limbu I et al. Vancomycin-induced immune
thrombocytopenia. N Engl J Med 2007; 356: 904–10.
2 Aster RH, Bougie DW. Drug-induced immune
thrombocytopenia. N Engl J Med 2007; 357: 580–7.
3 Rubinstein EE. History of quinolones and their side
effects. Chemotherapy 2001; 47 (Suppl. 3): 3–8; discussion
44–8.
4 Ball P. Quinolone generations: natural history or natural
selection? J Antimicrob Chemother 2000; 46 (Suppl. 3):
17–24.
5 Blum MMD, Graham DDJ, McCloskey CCA.
Temafloxacin syndrome: review of 95 cases. Clin Infect Dis
1994; 18: 946–50.
6 George JN, Raskob GE, Shah SR, Rizvi MA,
Hamilton SA, Osborne S et al. Drug-induced
thrombocytopenia: a systematic review of published
case reports. Ann Intern Med 1998; 1129 (11_Part_1):
886–90.
7 Kaufman DDW, Kelly JJP, Johannes CCB, Sandler AA,
Harmon DD, Stolley PPD et al. Acute thrombocytopenic
purpura in relation to the use of drugs. Blood 1993; 82:
2714–18.
8 George JN. Thrombotic Thrombocytopenic Purpura. N
Engl J Med 2006; 354: 1927–35.
9 Maguire RRB, Stroncek DDF, Gale EE, Yearlsey MM.
Hemolytic anemia and acute renal failure associated with
temafloxacin-dependent antibodies. Am J Hematol 1994;
46: 363–6.
10 Allan DSDS, Thompson CMCM, Barr RMRM,
Clark WFWF, Chin-Yee IHIH. Ciprofloxacin-associated
hemolytic-uremic syndrome. Ann Pharmacother 2002; 36:
1000–2.
11 Mouraux AA, Gille MM, Piéret FF, Declercq II.
Fulminant thrombotic thrombocytopenic purpura in the
course of ciprofloxacin therapy. Rev Neurol 2002; 158:
1115–17.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Fluoroquinolone-induced immune thrombocytopenia
12 Chichmanian RRM, Spreux AA, Bernard EE,
Garraffo RR, Fuzibet JJG. Thrombopenia due to
pefloxacin (Peflacine): dose-dependent toxicity? Thérapie
1992; 47: 419–21.
13 Teh CC, McKendrick MM. Ciprofloxacin-induced
thrombocytopenia. J Infect 1993; 27: 213–15.
14 Dutta TTK, Badhe BBA. Ciprofloxacin-induced bone
marrow depression. Postgrad Med J 1999; 75: 571–3.
15 Meyboom RRH. Thrombocytopenia induced by nalidixic
acid. Br Med J 1984; 289: 962.
16 Lehmann MM, Arnaud CC, Le Quellec AA,
Galley-Rand MM, Ciurana AAJ, Blayac JJP.
Thrombocytopenia probably induced by norfloxacin.
Apropos of a case. Thérapie 1991; 46: 410–11.
17 Lecomte I, Thioliere B, Azanowsky JM.
Pefloxacin-induced thrombocytopenia. One case report.
Rev Med Interne 1991; 12: 1.
18 Denis JJP, Martin CC, Gouin FF. Absence of recurrence
with ciprofloxacin therapy of thrombocytopenia
induced by pefloxacin. Ann Fr Anesth Reanim 1992; 11:
726.
19 Aydogdu II, Ozerol IIH, Tayfun EE, Harputluoglu MM.
Autoimmune haemolytic anaemia and thrombocytopenia
associated with ciprofloxacin. Clin Lab Haematol 1997; 19:
223.
20 Wensing JJW, Vlasveld LLT. Immune thrombocytopenia
attributed to norfloxacin. Ned Tijdschr Geneeskd 1997; 141:
1660–2.
21 Gales BBJ, Sulak LLB. Severe thrombocytopenia
associated with alatrofloxacin. Ann Pharmacother 2000;
34: 330–4.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
22 Takahama HH, Tazaki HH. Tosufloxacin tosilate-induced
thrombocytopenic purpura. J Dermatol 2007; 34: 465–7.
23 Starr JAJA, Ragucci KRKR. Thrombocytopenia associated
with intravenous ciprofloxacin. Pharmacotherapy 2005;
25: 1030–4.
24 Chamouard PP, Duclos BB, Welsch MM, Gold AA.
Severe reversible thrombopenia induced by norfloxacin.
Presse Med 1987; 16: 1978–9.
25 Kinoshita YY, Yamane TT, Kamimoto AA, Oku HH,
Iwata YY, Kobayashi TT et al. A case of
pseudothrombocytopenia during antibiotic
administration. Rinsho Byori 2004; 52: 120–3.
26 Campi PP, Pichler WJWJ. Quinolone hypersensitivity.
Curr Opin Allergy Clin Immunol 2003; 3: 275–81.
27 Aster RRH. Quinine sensitivity: a new cause of the
hemolytic uremic syndrome. Ann Intern Med 1993; 119:
243–4.
28 Crum NNF, Gable PP. Quinine-induced hemolytic-uremic
syndrome. South Med J 2000; 93: 726–8.
29 Gottschall JJL, Neahring BB, McFarland JJG, Wu GGG,
Weitekamp LLA, Aster RRH. Quinine-induced immune
thrombocytopenia with hemolytic uremic syndrome:
clinical and serological findings in nine patients and
review of literature. Am J Hematol 1994; 47: 283–9.
30 Reddy JC, Shuman MA, Aster RH. Quinine/
quinidine-induced thrombocytopenia: a great imitator.
Arch Intern Med 2004; 164: 218–20.
31 Bougie DW, Wilker PR, Aster RH. Patients with
quinine-induced immune thrombocytopenia have both
‘drug-dependent’ and ‘drug-specific’ antibodies. Blood
2006; 108: 922–7.
623
Internal Medicine Journal 39 (2009) 624–627
H I S TO RY I N M E D I C I N E
Medicinal use of leeches in the texts of ancient Greek, Roman
and early Byzantine writers
imj_1965
624..627
N. Papavramidou1 and H. Christopoulou-Aletra2
1
History of Medicine, Department of Anatomy, School of Medicine, Democritus University of Thrace, Alexandroupolis, and 2History of Medicine,
School of Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece
Key words
history of leech, medicinal leech, antiquity,
leech.
Correspondence
Helen Christopoulou-Aletra, 73 Nikis Avenue,
54622, Thessaloniki, Greece.
Email: [email protected]; ealetra@
med.auth.gr
Received 24 September 2008; accepted 10
October 2008.
doi:10.1111/j.1445-5994.2009.01965.x
Abstract
Blood-letting was a common therapeutic method in antiquity; many means
were used to draw blood, including the application of leeches. In this paper,
ancient Greek, Roman and Byzantine authors up to the 7th century AD were
studied, a research that provided us with references that may be divided into
two groups: those related to the medicinal use of leeches, and those related to
cases in which leeches were swallowed and had to be removed. In the first
group, detailed descriptions of the method of usage and of the diseases requiring leeching were found. In the second group, brief reference is made to the
problems caused by swallowing leeches, and to the methods used to expel
them from the human organism. The earliest references to the medicinal use
of leeches may be found in the writings of Theocritus (3rd century BC),
Nicander (2nd century BC) and Horace (1st century BC, while the phenomenon of swallowing a leech is first mentioned in one of the Epidaurian ‘iamata’
dating to the 4th century BC.
Introduction
Medicinal use of leeches
Blood-letting was one of the most common therapeutic
methods in Greek, Greco-Roman and Byzantine medicine. The methods used for this purpose were cupping,
venesection and application of leeches. Each method
was used in different cases, depending on the desirable
amount of blood to be extracted and the part of the body
affected. For the present paper, we studied texts from the
4th century BC to the 7th century AD.
Leeches appear in the Biblical Proverbs as always being
unsatisfied and craving for more blood.1 Aristotle, in his
De incessu animalium, refers – strictly from the standpoint
of natural history – to the way leeches move, and
compares them to worms.2 The ability of leeches to
suck blood is mentioned in the poetry of Theocritus.3
In Horace’s The art of poetry: to the Pisos, a leech is
described as an animal that would not abandon the skin,
unless satiated with blood; in the same way, a mad poet
can become tenaciously attached to anything or anyone.4
On the other hand, Dioscorides (1st century AD) lists
leeches among the poisonous animals in his book De
venenis eorumque praecautione et medicatione.5
The first use of leeches for medical purposes appears in
the wall of an Egyptian tomb (1567–1308 BC),6 but it
is only to Nicander that the earliest written evidence is
attributed.
In the 5th century AD, Caelius Aurelianus mentions
that Themison of Laodicea was the first to use leeches for
medical purposes.7 However, Nicander of Colophon
appears in the literature to be the actual first, when
writing that one should apply leeches after having been
bitten by a poisonous animal, so that they extract the
poisoned blood.8
Pliny the Elder believes that leeches relieve the body of
superfluous blood and open the pores of the skin. He
mentions, however, that once leeches have been used,
they have to be used again at the same period in each
succeeding year and that they only fall off when they lose
their grip through the weight of the blood; if they do not
fall off by virtue of their weight, they should be sprinkled
with salt. Pliny also mentions the mishap where a leech
leaves its ‘head’ buried in the flesh; such a case would
leave an incurable wound that could cause death as a
624
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Leeches in antiquity
result of the poisonous nature of leeches. Leeches,
according to Pliny, are appropriate for the treatment of
gout and for killing bugs, when used as an ingredient in
fumigators.9
Aretaeus of Cappadocia prescribes the use of leeches in
three different circumstances: in the case of ‘acute diseases of the liver’, in the case of ‘satyriasis’ and for the
treatment of ‘abdominal diseases’. In the first case, the
necessity for leeching derives from the physiology of
the liver, because the production of blood is attributed to
that organ, whence it is distributed to the entire body. If
the passageways are closed, the liver becomes inflamed
and bleeding should be induced.10 In the case of ‘satyriasis’, which was believed to be an inflammation of the
nerves of the pubic area causing genital erection accompanied by intense sexual desire, blood should be removed
with the aid of leeches.10 Finally, leeches are prescribed
for the treatment of ‘abdominal diseases’, caused by the
inability of the stomach to digest food.11
During the same period, Demosthenes Philalethes,
physician and member of the school founded by Herophilus,12 refers to the use of leeches for the treatment of
‘paralysis of the eye’.13
Archigenes (1st century AD) also refers to leeches,
information acquired through the physician Posidonius,
(1st century AD) and Aetius of Amida (6th century AD).
According to Archigenes, leeches should be used for the
treatment of ‘mania’, by applying them in a circle around
the entire head, especially on the bregma.13
A century later, Galen describes the medicinal use of
leeches in his treatise De hirundinibus, cucurbitula, incisione
et scarificatione. He first refers to their preparation before
use: leeches just found should be kept in a vase for a day
and fed with a little blood, in order to decrease their
‘poison’. Then, the site of the body where leeches are to
be placed should be cleaned with niter and scratched
with the nails, to increase blood circulation. The leeches
should first be placed in tepid water and cleaned of the
slime covering them with a sponge. Next, tepid oil was to
be poured on the body member. After leeches began
sucking, it was difficult to extract them, unless the doctor
put some ‘salt or ash’ on their ‘mouth’. After their
removal, cupping was to be used to extract the poison left
at the site from which the leeches drew blood. If blood
dribbled from the wound left by the leeches, cumin or
flour should be sprinkled on the site.14
The general indication for the use of leeches, described
in the pseudo-Galenic treatise Definitiones medicae, is an
excess of blood, especially when the physician believes
that it is the cause of a disease.15 Leeches are helpful in
cases of satyriasis or priapism,13 or nasal congestion. In
the latter case, leeches were to be applied to the tip of the
nose to extract the blood gathered there.13
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
In the same period (2nd century AD), another physician, Menemachus from Aphrodisias, wrote numerous
works no longer extant. We have his views on the use
of leeches via Oribasius (4th century AD). According to
Menemachus, leeches are applied on ailing parts or on
places adjacent to them. The site of application of leeches
should be free of fat, ‘because the fat stops their appetite’.
In order to extract the leeches from the affected bodily
part, Menemachus proposes the use of hot oil, and
renounces Galen’s use of salt. If the leeches are slow in
drawing blood, the physician should superficially scarificate the skin, because ‘when leeches taste blood, they
search for it even more’. The quantity of blood sucked
may be evaluated either by squeezing the leeches and
emptying the blood or by gathering the blood they
‘vomit’ after detachment from the body.16
The Emperor Flavius Claudius Julianus, also known as
Julian the Apostate, whose physician was Oribasius,12
refers to leeches and their ability to suck unhealthy
blood from the body and leave the healthy blood
untouched.17
At the same time, Antyllus, another surgeon, refers to
the medicinal use of leeches in his treatise On the means of
depletion. Again, none of his works has been preserved
and we know of his opinions through the writings of
Oribasius, Paulus of Aegina and Aetius of Amida. Antyllus suggests the use of leeches 1 day after they have been
found. After being stored and fed with some blood, they
are left for a period of time to move about, so as to
diminish their venom. Before leeches are applied to the
skin, the affected part should be rubbed with unrefined
soda and coated with the blood of animal or damp clay, or
scratched by the physician’s nails. ‘Thanks to these preparations, the leeches suck blood more promptly.’16 In order
for the leeches to keep sucking blood, the physician
should cut their tails with scissors. To extract them from
the skin, salt, ash or unrefined soda should be applied on
their mouth. Their venom should then be removed from
the affected part with a specific instrument, or by fumigation. If blood leaks from the wound left by the leeches’
bite, it should be sprinkled with ash from frankincense,
cumin or flour, and then be covered with linen soaked in
oil. In case extremities are involved, a bandage is used,
and if the bleeding has stopped on the second day, then
the area is washed. According to Antyllus, leeches do not
draw blood from the inner structures, only from superficial ones. Leeches are applied to those patients that are
afraid of scarification or on those body parts where
cupping cannot be applied, due to their small size or their
curvature. The leeches are removed from the body when
they have sucked half the blood initially scheduled for
removal. The physician should then allow blood to run
from the wound for a while.16
625
Papavramidou & Christopoulou-Aletra
Caelius Aurelianus, another famous physician, suggests the use of leeches in cases of headache or mania,
with their application on the head, and in cases of inflammation of the throat, with their application to the neck.18
Two centuries later, in the 6th century AD, Aetius of
Amida notes that leeches should be used in two cases: in
that of ‘catharsis’13 or in the case of an eye disease called
‘onychia’. ‘Catharsis’ is the restoration of the distorted
balance of bodily fluids, specifically blood. ‘Onychia’ is
described as a disease where pus accumulates from a deep
ulcer between the folds of the eye and forms a circle
around the iris. In this case, leeches should be applied to
the temples.13
Alexander of Tralles, at the same period, prescribes the
application of leeches to the head for the treatment of
‘melancholy’.19 Such application makes sense, bearing in
mind that in ancient times, ‘melancholy’ was thought to
be caused by an excess of black bile resulting in diseased
condition of either the brain or the entire bloodstream.20
In the latter case, extracting the diseased blood with the
aid of leeches could solve the problem.
Paulus Aegineta, a celebrated physician of the 7th
century AD, suggests the use of leeches in many cases. He
prescribes leeching in the case of ‘cephalea’, a permanent
pain of the head aggravated by noises, cries, bright light,
wine-drinking and intense smells. When ‘cephalea’
becomes chronic, the use of leeches is indicated.18 Paulus
also suggests leeches for two eye diseases: ‘amaurosis’
and ‘ophthalmy’. ‘Amaurosis’ was defined as ‘a complete
impediment of the sight without any apparent affection
about the eye, without any sensible cause’. For its treatment, the application of leeches to the temples was necessary.18 ‘Ophthalmy’ was an inflammation of one or
more membranes of the eye, especially the conjunctiva.21
So, if the cause of this disease is a congestion of the
humours accumulated in the head, leeches should be
applied to the forehead, near the affected eye.18 Another
disease requiring leeching to which Paulus Aegineta
refers is ‘synanche’, which appears to be an inflammation
of the throat. In this case, leeches should be applied to the
chin and neck of the patient.18 Finally, leeching is prescribed for headaches accompanied by fever, and in the
case of mania, in which cases they are applied especially
to the head.18
Discussion
The Greek word ‘bdlla’ derives from the verb ‘bdllw’,
‘to suck’,22 so one of the abilities of leeches is indicated by
the animal’s name itself. In his Natural History, Pliny the
Elder refers to leeches by their Latin name ‘sanguisuga’
(sanguis = blood, ‘suctus’ = to suck).9 Here as well, the
ability of leeches is clearly denoted. A new name was
626
attributed to leeches by Carl von Linné (1707–1778):
‘Hirudo medicinalis’; here, the term used referred directly
to the medicinal use of leeches.6 Their current English
name derives from the old English word ‘leace’, meaning
doctor.23 The successive renaming of leeches through the
centuries proves not only their long-standing and important role in medicine but also their effectiveness. Their
medicinal use lasted for centuries, reaching its peak
during the 19th century, when great numbers were
imported from various countries to cover medical needs;
in France, for example, 30 million leeches per year were
imported.23
Galen introduces the use of leeches, which had been
altogether ignored by the Hippocratic physicians. He
agrees with Dioscorides, Pliny and Antyllus about the
‘poisonous’ nature of leeches, for which Daremberg
attempts to give an explanation, reckoning that the application of leeches to a diseased person made it easy to
transmit the infected blood to a relatively healthier person.16 Thus, the presence of ‘poison’ gave a satisfactory
interpretation. The point Daremberg makes, however, is
not proved, as direct re-use of the same leech is not
mentioned by the authors and was probably impossible,
given that leeches are fed from once every 6 months to
once a year.24 Today, things may be explained in full
scientific way: a bacterium, the Aeromonas hydrophila,
thrives in the leech’s gut and has been shown to be
responsible for causing infection after its use.25 Galen was
wise enough to advise cupping with scarification to
extract the ‘poison’ and Antyllus also advised allowing
blood to run from the wound after the application of
leeches.
Until the beginning of the second decade of the 20th
century, physicians supported the use of leeches.23 Today,
they are used experimentally for venous congestion, for
the anticoagulant substance of their saliva, and the antihistaminic vasodilator property that promotes slow
bleeding from the area to which a leech is applied. It is
also considered that leeches can be used for several kinds
of thromboses in lieu of heparin.23 Thus, contemporary
research on the properties found in leeches has explained
the haemorrhage after the suction of blood, because we
now know the anticoagulant properties of the leeches’
saliva.
Leeches were used in all times for their decongestant
properties and are indeed still used for this purpose. They
promote better local haemostasis, they prevent necrosis
or apoptosis of an implanted structure, they induce anticoagulant treatment (with the secretion of hirudin) and
they secrete anaesthetic, an ‘antihistamine-like vasodilation, hyaluronidase that promotes the spread of saliva
into the bitten wound’.24 They are also used in reconstructive surgery for the restoration of microvascular
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Leeches in antiquity
congestion.25 It appears that the positive effects of this
ancient remedy may now be explained through scientific
methods, promising potentially even more uses of this
admirable creature in medicine.
References
1 Bible. The Old Testament; Proverbs 30: 15. Athens: The
Greek Bible Society, 1997; 990.
2 Aristotle. De Incessu Animalium (Farquharson ASL, trans.).
Oxford: Clarendon Press; 1912; 9.
3 Theocritus. The sorceress. In: Hunter R, trans. Theocritus.
Idylls. London: Oxford University Press; 2002; 8.
4 Horace. The art of poetry: to the Pisos. In: Smart C,
trans. Horace. The Works of Horace. New York: Harper &
Brothers; 1863; v. 476.
5 Dioscorides. De venenis eorumque praecautione et
medicatione. In: Sprengel K, ed. Pedanii Dioscoridis
Anazarbei Perí dhlhthríwn, iobólwn kai euporístwn
Spuria. Lipsia: Car. Cnoblochii; 1830; 14.
6 Whitaker IS, Rao J, Izadi D, Butler PE. Hirudo
medicinalis: ancient origins of leeches, and trends in the
use of medicinal leeches throughout history. Br J Oral
Maxillofac Surg 2004; 42: 133–7.
7 Caelius Aurelianus’. On Chronic Diseases. In: Drabkin IE,
ed. and trans. Caelius Aurelianus: On Acute and on Chronic
Diseases. Chicago, IL: The University of Chicago Press;
1950; 286.
8 Schneider O. Nicandrea, Theriaca et Alexipharmaca. Lipsia:
BG. Teubneri; 1856; 271.
9 Pliny the Elder. Natural History. Cambridge, MA: Harvard
University Press; 1989; 539.
10 Aretaeus. De Curatione Acutorum Morborum. Athens:
Kaktos; 1997; 142, 165.
11 Aretaeus. De Curatione Diuturnorum Morborum. Athens:
Kaktos; 1997; 233.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
12 Smith W. Dictionary of Greek and Roman Antiquities, Vol. 1.
London: C. Littré and J. Brown; 1870; 991, 644.
13 Aetius. Librorum Medicinalium. Venise: Aldus Manutius;
1534; 134r, 100v, 122v, 127v.
14 Galen. De hirundinibus, cucurbitula, incisione et
scarificatione. In: Galen. Opera Omnia, Vol. 11. Lipsia: Car.
Cnoblochii; 1826; 317.
15 Galen. Definitiones medicae. In Galen. Opera Omnia, Vol.
19. Lipsia: Car. Cnoblochii; 1830; 458.
16 Daremberg Ch. Oeuvres d’Oribase, Vol. 2. Paris:
Imprimeries Impériale; 1884; 69–73, 790–1.
17 Flavius Claudius Julianus. Contra Galileos. In:
Neumann CJ, ed. Juliani imperatoris librorum contra
Christianos quae supersunt. Leipzig: Teubner; 1880; 198.
18 Adams F. The Seven Books of Paulus Aegineta, Vol. 1.
London: Sydenham Society; 1844; 302, 357–8, 385, 410,
421, 464–5.
19 Pushmann T. Alexander von Tralles, Vol. 1. Wien: W.
Braumuller; 1878; 593–603.
20 Papavramidou N. The notions of ‘psyche’ and ‘mental
illness’ in the Hippocratic, Aristotelian and Galenic texts
[PhD thesis]. Thessaloniki: Aristotle University of
Thessaloniki, Greece; 2006; 161.
21 Encyclopédie méthodique, Médecine, Vol. 8. Paris: Vve
Agasse; 1824; 143.
22 Liddell HG, Scott R, Jones HS, McKenzie R. A
Greek-English lexicon. Cambridge: Clarendon Press; 1992;
312.
23 Upshaw J, O’Leary JP. The medical leech: past and
present. Amer Surgeon 2000; 66: 313–14.
24 Nowak G, Schror K. Hirudin – the long and stony way
from an anticoagulant peptide in the saliva of medicinal
leech to a recombinant drug and beyond. Thromb Haemost
2007; 87: 116–19.
25 O’Hara M. Leeching: a modern use from an ancient
remedy. Am J Nurs 1988; 88: 1656–58.
627
Internal Medicine Journal 39 (2009) 628
I M AG E S I N M E D I C I N E
Complications of thoracentesis
imj_1993
A 58-year-old female patient presented with shortness of
breath. Chest radiograph revealed a large right pleural
effusion (Fig. 1). She underwent thoracentesis with the
removal of 2.5 L of fluid. Post procedure, the patient
developed mild respiratory distress. A chest computed
tomography scan was done (Fig. 2) that showed small
right pneumothorax and diffuse right sided ground glass
opacification consistent with re-expansion pulmonary
oedema (REPE). The patient’s symptoms improved
with supportive measures. Cytological examination of
the fluid revealed adenocarcinoma, and later on, she
underwent pleurodesis.
Thoracentesis is a useful and safe procedure. Pneumothorax following thoracentesis is reported in 12% of
cases.1 It is usually managed conservatively with oxygen
supplementation; however, chest tube drainage may
be required. REPE is another rare complication that is
reported after thoracentesis of a large pleural effusion or
chest tube drainage of a large pneumothorax. REPE is
thought to be secondary to increased vascular permeability when the lung suddenly re-expands.2 It is more likely
to develop with prolonged duration of the pneumothorax
or pleural effusion and the rapidity of lung re-expansion.3
The manifestations of REPE range from asymptomatic
radiological phenomenon to acute respiratory failure that
develops soon after the procedure and may require
mechanical ventilation.3
Figure 1 A posterio-anterior view of the chest radiograph showing large
right pleural effusion.
628
628
It is recommended in the case of large and prolonged
pleural effusion to drain the fluid slowly and not to
exceed 1.5 L at a time.4 In the case of large pneumothorax, it is advisable to connect the chest tube to a
water seal without negative pressure.4
Received 16 February 2009; accepted 9 March 2009.
doi:10.1111/j.1445-5994.2009.01993.x
A. O. Soubani and M. Valdivieso
Karmanos Cancer Center and Wayne State University
School of Medicine
Detroit, Michigan, USA
References
1 Collins TR, Sahn SA. Thoracocentesis: clinical value,
complications, technical problems, and patient experience.
Chest 1987; 91: 817–22.
2 Matsuura Y, Nomimura T, Murakami H, Matsushima T,
Kakehashi M, Kajihara H. Clinical analysis of reexpansion
pulmonary edema. Chest 1991; 100: 1562–6.
3 Sherman SC. Reexpansion pulmonary edema: a case
report and review of the current literature. J Emerg Med
2003; 24: 23–7.
4 Echevarria C, Twomey D, Dunning J, Chanda B. Does
re-expansion pulmonary oedema exist? Interact Cardiovasc
Thorac Surg 2008; 7: 485–9.
Figure 2 A computed tomography image of the chest following thoracentesis showing small right pneumothorax and diffuse right sided
ground glass infiltrate consistent with re-expansion pulmonary edema.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Internal Medicine Journal 39 (2009) 629–632
L E T T E R S TO T H E E D I TO R
Clinical-scientific notes
imj_1999
629..632
Pulmonary toxicity associated with
infliximab therapy for ulcerative colitis
Monoclonal anti-tumour necrosis factor-a (TNF-a) antibody (infliximab) is used in the management of rheumatological disorders and inflammatory bowel disease.1
Although it is generally a well-tolerated drug, there is
concern regarding its pulmonary toxicity.2–5 In this case
report we describe interstitial pneumonitis following
infliximab infusion in a patient with ulcerative colitis.
A 76-year-old man with intractable ulcerative colitis
received one dose (5 mg/kg) of infliximab. Four weeks
following infliximab infusion he developed severe dyspnoea with hypoxaemic respiratory failure. His past
medical history was unremarkable other than wellcontrolled essential hypertension and he was on therapeutic anticoagulation for previously diagnosed deep
vein thrombosis. He was a reformed smoker with a 30
pack year of smoking history with no prior respiratory
symptoms. He had no significant history of occupational
or environmental exposures relevant to lung disease.
Respiratory examination revealed faint bilateral endinspiratory crackles. The rest of the physical examination
was unremarkable; in particular there were no features
of cardiac failure.
Investigations showed moderate leucocytosis with a
white blood cell count of 14.9 ¥ 109/L and haemoglobin
of 120 g/L. Electrolytes and liver function tests were
normal. C-reactive protein was elevated at 54 mg/L.
Blood and sputum cultures were negative. Detailed
screening for bacterial, viral, including atypical microorganisms, such as Mycoplasma, Chlamydia and
Legionella were negative. The arterial blood gases
on 2 L/min of supplemental oxygen showed PaO2 of
62 mmHg and PaCO2 of 34 mmHg. Immunology screening showed positive anti-nuclear antibody with a titre of
160 (normal <80) and weekly positive anti-neutrophil
cytoplasmic antibody. Chest X-ray showed bilateral widespread interstitial opacities. Echocardiogram showed
moderate aortic regurgitation with preserved left ventricular function. Despite treatment with empirical broad
spectrum antibiotics he showed no clinical improvement.
A high-resolution computed tomography (HRCT) of the
chest showed bilateral diffuse reticular opacities predominantly in the peripheral upper lung fields (Fig. 1), consistent with widespread interstitial pneumonitis/early
fibrosis. This was in contrast to a previous chest computed tomography scan few weeks prior to infliximab
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
infusion (Fig. 2). Bronchoscopic examination was
normal and broncho-alveolar lavage showed no evidence
of bacterial, viral, fungal or mycobacterial infection.
Although infliximab is known to cause interstitial
pneumonitis, it is only rarely reported in the literature.1–5
Villeneuve et al.3 in their review reported seven cases.
Most reported cases have been among patients with
rheumatoid arthritis.2–5 It has been speculated that
interstitial lung disease could be potentiated following
infliximab in the background of either methotrexate or
pre-existing interstitial lung disease secondary to rheumatoid arthritis.2,3,5 The mechanism of interstitial pneumonitis and infliximab is not entirely clear. However, it is
postulated that interaction of TNF-a with interleukin1 or
interferon-g may modulate fibroblast proliferation. Inhibition of TNF-a by infliximab may result in an increased
pro-inflammatory effect of these cytokines.2,3
In contrary to previously documented cases, our
patient had neither pre-existing lung fibrosis nor concomitant prior administration of pulmonary toxic drugs.
The temporal relationship between infliximab infusion
and striking change on the HRCT along with lack of
alternative diagnosis supports the view that infliximab
may be the primary cause of potentially fatal pulmonary
Figure 1 High-resolution computed tomography showing bilateral
diffuse pulmonary fibrosis following infliximab infusion.
629
Letters to the Editor
4 Kramer N, Chuzhin Y, Kaufman L, Ritter J,
Rosenstein ED. Methotrexate pneumonitis after initiation
of Infliximab therapy for rheumatoid arthritis. Arthritis
Rheum 2002; 47: 670–1.
5 Courtney PA, Alderdice J, Whitehead EM. Comment on
methotrexate pneumonitis after initiation of Infliximab
therapy for rheumatoid arthritis. Arthritis Rheum 2003;
49: 617.
General correspondence
Tako-tsubo cardiomyopathy after
observing anaphylaxis
Figure 2 Prior to infliximab infusion showing absence of reticular interstitial opacities.
toxicity. Further studies are warranted to assess the
incidence and monitoring strategies.
Acknowledgement
We thank Dr Matthew P Doogue, Department of Clinical
Pharmacology, for his expert opinion and advice.
Received 31 March 2008; accepted 22 April 2008.
doi:10.1111/j.1445-5994.2009.001999.x
S. S. Heraganahally,1 V. Au,2 S. Kondru,2 S. Edwards,3
J. J. Bowden1 and D. Sajkov1
3
Departments of 1Respiratory Medicine, 2Medical Imaging and
Gastroenterology, Flinders Medical Centre and Flinders University
Adelaide, South Australia, Australia
References
1 Ljung T, Karlén P, Schmidt D, Hellström PM, Lapidus A,
Janczewska I et al. Infliximab in inflammatory bowel
disease: clinical outcome in a population based cohort
from Stockholm county. Gut 2004; 53: 849–53.
2 Ostor AJ, Chilvers ER, Somerville MF, Lim AY, Lane SE,
Crisp AJ et al. Pulmonary complications of Infliximab
therapy in patients with rheumatoid arthritis. J
Rheumatol 2006; 33: 622–8.
3 Villeneuve E, St-Pierre A, Haraoui B. Interstitial
pneumonitis associated with Infliximab therapy.
Rheumatology 2006; 33: 1189–93.
630
We read with interest the recent article by Zubrinich
et al.1 which highlighted the association between exogenous catecholamines and tako-tsubo cardiomyopathy,
a phenomenon characterized by stress-related, transient,
left ventricular dysfunction in the absence of angiographically apparent coronary artery disease.
A wide range of possible triggers for tako-tsubo, including anaphylaxis, has been described.2,3 The pathophysiology may be related to endogenous catecholamine
toxicity and the report from Zubrinich et al. suggests that
exogenous catecholamines could also play a role.
We would like to extend this report to suggest that
simply observing and assisting in the treatment of
an anaphylactic episode can also trigger tako-tsubo
cardiomyopathy.
A 72-year-old woman with a past history of coronary
artery stenting, and previously documented normal left
ventricular function, was present when her daughter
suffered an anaphylactic reaction to seafood at their
country property. The daughter developed severe urticaria and hypotension. To expedite treatment, the
mother drove her daughter to meet the local ambulance
en route. After the rendezvous beside the Hume
freeway, the daughter’s condition was stabilized and
adrenaline administered. During these events the
mother developed chest pain and dyspnoea. Both
women were admitted to the local hospital. The daughter made an uneventful recovery. The mother was
noted to have new anterolateral T wave inversion on
her electrocardiograph and an elevated troponin of
11.0 mg/mL (normal < 0.04 mg/mL).
Coronary angiography showed wide patency of the
stented and non-stented vessels but severe left ventricular impairment because of anterior and apical akinesis,
not localized to a single epicardial coronary artery
distribution. Subsequent echocardiography performed
2 weeks after the event showed that her left ventricular
function had returned to normal.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
Letters to the Editor
Although the administration of adrenaline may trigger
tako-tsubo cardiomyopathy, our case indicates that even
a witness of anaphylaxis and adrenaline administration
may be at risk of the syndrome described by Zubrinich
et al.
Received 7 January 2009; accepted 15 January 2009.
doi:10.1111/j.1445-5994.2009.01926.x
C. Jellis, A. Hunter, R. Whitbourn and A. MacIsaac
Department of Cardiology, St Vincent’s Hospital,
Melbourne, Victoria, Australia
References
1 Zubrinich CM, Omar Farouque HM, Rochford SE,
Sutherland MF. Tako-tsubo-like cardiomyopathy after
epipen administration. Intern Med J 2008; 38: 862–5.
2 Connelly KA, MacIsaac AI, Jelinek VM. Stress, myocardial
infarction, and the ‘tako-tsubo’ phenomenon. Heart 2004;
90: e52.
3 Vultaggio A, Matucci A, Del Pace S, Simonetti I,
Parronchi P, Enrico M et al. Tako-Tsubo-like syndrome
during anaphylactic reaction. Eur J Heart Fail 2007; 9:
209–11.
Does alcohol play a role in QT
prolongation?
Alcohol consumption can add to the risks of many prescribed medicines. We audited clinical files of methadone
maintenance patients who had ECG recordings to
ascertain if alcohol self-medication may potentiate QTc
prolongation.
Methadone has held a long-standing role in the costeffective management of opiate addiction.1–3 However, a
recent awareness of its propensity to prolong the ratecorrected QT interval in a proportion of those treated has
generated both clinician wariness and debate.2,4–9 A possible molecular mechanism has now been uncovered.10
Some studies have demonstrated a moderate dose
relationship.2,9,11 Consequently, recommendations for
risk management and high-dose caution have been published,4,7,12 although other researchers have shown no
significant QTc–dose correlation.13
Predisposing factors, such as hypokalaemia, intercurrent illness, HIV infection, dehydration, cytochrome
P450 inhibitors and QT prolonging medications,12,14,15
explain the existence of QT prolongation in most reports
in the literature. There is no reason to expect increased
prevalence of familial QT syndromes among addiction
patients, but some people exhibit QTc prolongation while
taking relatively low methadone doses.16 There have also
been reports of significant QTc fluctuations over succes© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians
sive electrocardiogram (ECG) recordings for an individual, of paradoxical QTc increases when methadone
dose is reduced and of sudden deaths on methadone
despite prior normal ECG.16 In addition, some patients
live for years taking methadone and having significant
QTc prolongation without incident, in the absence of
other risk factors.17
Consequently, there is now a heightened interest in
monitoring the QTc during methadone maintenance,4,7
but there is no agreement over the dose at which the QT
prolongation risk increases, optimal frequency for ECG
screening,4 alternative therapy options18 and implications17 as treatment services seek to better understand
individual patient risk and to balance optimal control
of opiate addiction against the unknown risks of QT
prolongation.
There may yet be additional clinical risk factors that
operate in this particular setting. Patients with addiction
commonly self-medicate on alcohol and drugs (both licit
and illicit). This contributes to patient deaths and adds
therapeutic risk. An elegant case–control study19 demonstrated increased QT variability in patients during acute
alcohol intoxication, the QT variability correlating with
severity of the withdrawal symptoms. This could throw
new light on anecdotal reports of sudden ‘methadone’
deaths where alcohol abuse was also a possible contributing factor.20
As part of a clinical audit our own service conducted
a review of the current methadone patients who had
ECG tracings. The methodology and main findings are
described in Hyslop’s study.16 At the time, 71 ECGs
tracings were available for 60 clients, 32 of which were
pre-treatment recordings and 39 taken on methadone
maintenance. Linear regression analysis revealed a
moderate but significant dose-dependent relationship
and by extrapolation we estimated a 3.6-ms increase in
QTc occurs for every 10 mg methadone dose increase.16
Alcohol intake, more than 10 standard drinks per week,
was the only significant difference between the group of
28 normal interval ECGs and the 11 ECGs showing QTc
prolongation (risk ratio (RR) 3.5, 95% confidence interval (CI) 1.3–8.9) other than the prescribed methadone
dose itself. Some patients had had multiple ECGs, but the
result still held on analysis of just one ECG per individual
(RR 5.1, 95% CI 1.1–24.9). These patients were neither
intoxicated nor in withdrawal at the time of the ECG and
had no QTc prolongation risks other than prescribed
methadone.
This work appears to be the first of its kind. Further
studies are required for validation. Possible differences
with other study populations include prevalence of
genetic susceptibility to effects of alcohol or methadone
and confounding patterns of alcohol or methadone
631
Letters to the Editor
consumption. In New Zealand, where a black market
for methadone is prominent (greater than heroin or
cocaine), patients can readily supplement their prescribed methadone with illicitly purchased doses, and
new patients presenting for methadone maintenance are
rarely methadone naïve.16
Further research is required to clarify the nature of this
alcohol–QT link, including ethanol blood level correlation. Meantime, we recommend that alcohol selfmedication be considered a potential compounding factor
in QT prolongation. Wise clinical practice, mindful of the
resulting and unpredictable polypharmacy, is one of the
challenges of Addiction Medicine.
Acknowledgements
Our grateful appreciation to Mrs Jeh Sie Chan who
assisted in preparation of the manuscript. Thanks also to
Dr Andrew Byrne for his external review.
Received 1 February 2009; accepted 3 February 2009.
doi:10.1111/j.1445-5994.2009.01964.x
1
H. J. Moriarty, T. P. Flewett2 and B. A. Hyslop3
1
Department of Primary Health Care and General Practice,
University of Otago, 2CADS Capital Coast DHB, Wellington, and
3
MidCentral DHB, Palmerston North, New Zealand
Funding: This study was funded by University of Otago Research
Committee.
References
1 Krantz MJ, Lewkowiez L, Hays H, Woodroffe MA,
Robertson AD, Mehler PS. Torsade de pointes associated
with very-high-dose methadone. Ann Intern Med 2002;
137: 501–4.
2 Ehret GB, Voide C, Gex-Fabry M, Chabert J, Shah D,
Broers B et al. Drug-induced long QT syndrome in
injection drug users receiving methadone: high
frequency in hospitalized patients and risk factors. Arch
Intern Med 2006; 166: 1280–7.
3 Johnson R, Chutuape M, Strain E, Walsh S, Stitzer M,
Bigelow G. A comparison of levomethadyl acetate,
buprenorphine, and methadone for opioid dependence.
N Engl J Med 2000; 343: 1290–7.
4 Cruciani R. Methadone: to ECG or Not to ECG . . . that is
still the question. J Pain Symptom Manage 2008; 36:
545–52.
632
5 Martell BA, Arnsten JH, Ray B, Gourevitch MN. The
impact of methadone induction on cardiac conduction in
opiate users. Ann Intern Med 2003; 139: 154–5.
6 Martell BA, Arnsten JH, Krantz MJ, Gourevitch MN.
Impact of methadone treatment on cardiac repolarization
and conduction in opioid users. Am J Cardiol 2005; 95:
915–18.
7 Krantz MJ, Mehler PS. QTc prolongation: methadone’s
efficacy-safety paradox. Lancet 2006; 368: 556–7.
8 Piguet V, Desmeules J, Ehret G, Stoller R, Dayer P. QT
interval prolongation in patients on methadone with
concomitant drugs. J Clin Psychopharmacol 2004; 24:
446–8.
9 Krantz MJ, Kutinsky IB, Robertson AD, Mehler PS.
Dose-related effects of methadone on QT prolongation in
a series of patients with torsade de pointes.
Pharmacotherapy 2003; 23: 802–5.
10 Katchman AN, McGroary KA, Kilborn MJ, Kornick GA,
Manfredi PL, Woosley RL et al. Influence of opioid
agonists on cardiac human ether-a-go-go-related gene K+
currents. J Pharmacol Exp Ther 2002; 303: 688–94.
11 Cruciani RA, Sekine R, Homel P, Lussier D, Yap Y,
Suzuki Y et al. Measurement of QTc in patients receiving
chronic methadone therapy.
J Pain Symptom Manage 2005; 29: 385–91.
12 Al-Khatib SM, LaPointe NMA, Kramer JM, Califf RM.
What clinicians should know about the QT interval.
JAMA 2003; 289: 2120–7.
13 Maremmani I, Pacini M, Cesaroni C, Lovrecic M,
Perugi G, Tagliamonte A. QTc interval prolongation in
patients on long-term methadone maintenance therapy.
Eur Addict Res 2005; 11: 44–9.
14 Roden DM. Drug-induced prolongation of the QT
interval. N Engl J Med 2004; 350: 1013–22.
15 Vorchheimer DA. What is QT interval prolongation?
J Fam Pract 2005; June(Suppl): S4–7.
16 Hyslop B. Prolongation of the QT interval during
methadone use: how important is the dose? NZ Med Stud
J 2007; 6: 4–7.
17 Byrne A, Stimmel B. Methadone and QTc prolongation
[letter]. Lancet 2007; 369: 366.
18 Wedam EF, Bigelow GE, Johnson RE, Nuzzo PA,
Haigney MCP. QT interval effects of methadone,
Levomethadyl and Buprenorphine in a randomised trial.
Arch Intern Med 2007; 167: 2469–73.
19 Bär KJ, Boettgerb MK, Koschkea M, Boettgera S,
Groteluschena M, Vossc A et al. Increased QT interval
variability index in acute alcohol withdrawal. Drug
Alcohol Depend 2007; 89: 259–66.
20 Pimental L, Mayo D. Chronic methadone therapy
complicated by torsades de pointes; a case report. J Emerg
Med 2008; 34: 287–90.
© 2009 The Authors
Journal compilation © 2009 Royal Australasian College of Physicians